Merge branch 'main' into sd3-t5

* Add vae_roundtrip.py example

* Add cuda support to vae_roundtrip

* Move vae_roundtrip.py into research_projects/vae

* Fix channel scaling in vae roundrip and also support taesd.

* Apply ruff --fix for CI gatekeep check

---------

Co-authored-by: Álvaro Somoza <asomoza@users.noreply.github.com>

.github/ISSUE_TEMPLATE/bug-report.yml

@@ -57,50 +57,54 @@ body:
       description: |
         Your issue will be replied to more quickly if you can figure out the right person to tag with @.
         If you know how to use git blame, that is the easiest way, otherwise, here is a rough guide of **who to tag**.
-        
+
         All issues are read by one of the core maintainers, so if you don't know who to tag, just leave this blank and
         a core maintainer will ping the right person.
-        
+
         Please tag a maximum of 2 people.
 
-        Questions on DiffusionPipeline (Saving, Loading, From pretrained, ...):
+        Questions on DiffusionPipeline (Saving, Loading, From pretrained, ...): @sayakpaul @DN6
 
         Questions on pipelines:
-        - Stable Diffusion @yiyixuxu @DN6 @sayakpaul 
-        - Stable Diffusion XL @yiyixuxu @sayakpaul @DN6 
-        - Kandinsky @yiyixuxu 
-        - ControlNet @sayakpaul @yiyixuxu @DN6 
-        - T2I Adapter @sayakpaul @yiyixuxu @DN6 
-        - IF @DN6 
-        - Text-to-Video / Video-to-Video @DN6 @sayakpaul 
-        - Wuerstchen @DN6 
+        - Stable Diffusion @yiyixuxu @asomoza
+        - Stable Diffusion XL @yiyixuxu @sayakpaul @DN6
+        - Stable Diffusion 3: @yiyixuxu @sayakpaul @DN6 @asomoza
+        - Kandinsky @yiyixuxu
+        - ControlNet @sayakpaul @yiyixuxu @DN6
+        - T2I Adapter @sayakpaul @yiyixuxu @DN6
+        - IF @DN6
+        - Text-to-Video / Video-to-Video @DN6 @sayakpaul
+        - Wuerstchen @DN6
         - Other: @yiyixuxu @DN6
+        - Improving generation quality: @asomoza
 
         Questions on models:
-        - UNet @DN6 @yiyixuxu @sayakpaul 
-        - VAE @sayakpaul @DN6 @yiyixuxu 
-        - Transformers/Attention @DN6 @yiyixuxu @sayakpaul @DN6 
+        - UNet @DN6 @yiyixuxu @sayakpaul
+        - VAE @sayakpaul @DN6 @yiyixuxu
+        - Transformers/Attention @DN6 @yiyixuxu @sayakpaul
 
-        Questions on Schedulers: @yiyixuxu 
+        Questions on single file checkpoints: @DN6
 
-        Questions on LoRA: @sayakpaul 
+        Questions on Schedulers: @yiyixuxu
 
-        Questions on Textual Inversion: @sayakpaul 
+        Questions on LoRA: @sayakpaul
 
-        Questions on Training: 
-        - DreamBooth @sayakpaul 
-        - Text-to-Image Fine-tuning @sayakpaul 
-        - Textual Inversion @sayakpaul 
-        - ControlNet @sayakpaul 
+        Questions on Textual Inversion: @sayakpaul
 
-        Questions on Tests: @DN6 @sayakpaul @yiyixuxu 
+        Questions on Training:
+        - DreamBooth @sayakpaul
+        - Text-to-Image Fine-tuning @sayakpaul
+        - Textual Inversion @sayakpaul
+        - ControlNet @sayakpaul
+
+        Questions on Tests: @DN6 @sayakpaul @yiyixuxu
 
         Questions on Documentation: @stevhliu
 
         Questions on JAX- and MPS-related things: @pcuenca
 
-        Questions on audio pipelines: @DN6 
-        
+        Questions on audio pipelines: @sanchit-gandhi
+
+
 
-        
       placeholder: "@Username ..."

.github/PULL_REQUEST_TEMPLATE.md

@@ -38,9 +38,9 @@ members/contributors who may be interested in your PR.
 
 Core library:
 
-- Schedulers: @yiyixuxu 
-- Pipelines:  @sayakpaul @yiyixuxu @DN6
-- Training examples: @sayakpaul 
+- Schedulers: @yiyixuxu
+- Pipelines and pipeline callbacks: @yiyixuxu and @asomoza
+- Training examples: @sayakpaul
 - Docs: @stevhliu and @sayakpaul
 - JAX and MPS: @pcuenca
 - Audio: @sanchit-gandhi
@@ -48,7 +48,7 @@ Core library:
 
 Integrations:
 
-- deepspeed: HF Trainer/Accelerate: @pacman100
+- deepspeed: HF Trainer/Accelerate: @SunMarc
 
 HF projects:
 

.github/workflows/benchmark.yml

@@ -13,13 +13,15 @@ env:
 
 jobs:
   torch_pipelines_cuda_benchmark_tests:
+    env: 
+      SLACK_WEBHOOK_URL: ${{ secrets.SLACK_WEBHOOK_URL_BENCHMARK }}
     name: Torch Core Pipelines CUDA Benchmarking Tests
     strategy:
       fail-fast: false
       max-parallel: 1
     runs-on: [single-gpu, nvidia-gpu, a10, ci]
     container:
-      image: diffusers/diffusers-pytorch-cuda
+      image: diffusers/diffusers-pytorch-compile-cuda
       options: --shm-size "16gb" --ipc host -v /mnt/hf_cache:/mnt/cache/ --gpus 0
     steps:
       - name: Checkout diffusers
@@ -50,4 +52,14 @@ jobs:
         uses: actions/upload-artifact@v2
         with:
           name: benchmark_test_reports
-          path: benchmarks/benchmark_outputs
+          path: benchmarks/benchmark_outputs
+
+      - name: Report success status
+        if: ${{ success() }}
+        run: |
+          pip install requests && python utils/notify_benchmarking_status.py --status=success
+        
+      - name: Report failure status
+        if: ${{ failure() }}
+        run: |
+          pip install requests && python utils/notify_benchmarking_status.py --status=failure

.github/workflows/mirror_community_pipeline.yml

@@ -22,6 +22,9 @@ on:
 
 jobs:
   mirror_community_pipeline:
+    env:
+      SLACK_WEBHOOK_URL: ${{ secrets.SLACK_WEBHOOK_URL_COMMUNITY_MIRROR }}
+    
     runs-on: ubuntu-latest
     steps:
       # Checkout to correct ref
@@ -36,7 +39,7 @@ jobs:
       #     If ref is 'refs/heads/main' => set 'main'
       #     Else it must be a tag => set {tag}
       - name: Set checkout_ref and path_in_repo
-        run: | 
+        run: |
           if [ "${{ github.event_name }}" == "workflow_dispatch" ]; then
             if [ -z "${{ github.event.inputs.ref }}" ]; then
               echo "Error: Missing ref input"
@@ -54,7 +57,7 @@ jobs:
           else
             # e.g. refs/tags/v0.28.1 -> v0.28.1
             echo "CHECKOUT_REF=${{ github.ref }}" >> $GITHUB_ENV
-            echo "PATH_IN_REPO=${${{ github.ref }}#refs/tags/}" >> $GITHUB_ENV
+            echo "PATH_IN_REPO=$(echo ${{ github.ref }} | sed 's/^refs\/tags\///')" >> $GITHUB_ENV
           fi
       - name: Print env vars
         run: |
@@ -86,4 +89,14 @@ jobs:
         run: huggingface-cli upload diffusers/community-pipelines-mirror ./examples/community ${PATH_IN_REPO} --repo-type dataset
         env:
             PATH_IN_REPO: ${{ env.PATH_IN_REPO }}
-            HF_TOKEN: ${{ secrets.HF_TOKEN_MIRROR_COMMUNITY_PIPELINES }}
+            HF_TOKEN: ${{ secrets.HF_TOKEN_MIRROR_COMMUNITY_PIPELINES }}
+
+      - name: Report success status
+        if: ${{ success() }}
+        run: |
+          pip install requests && python utils/notify_community_pipelines_mirror.py --status=success
+      
+      - name: Report failure status
+        if: ${{ failure() }}
+        run: |
+          pip install requests && python utils/notify_community_pipelines_mirror.py --status=failure

.github/workflows/notify_slack_about_release.yml

@@ -11,12 +11,12 @@ jobs:
 
     steps:
     - uses: actions/checkout@v3
-    
+
     - name: Setup Python
       uses: actions/setup-python@v4
       with:
         python-version: '3.8'
-    
+
     - name: Notify Slack about the release
       env:
         SLACK_WEBHOOK_URL: ${{ secrets.SLACK_WEBHOOK_URL }}

.github/workflows/pr_dependency_test.yml

@@ -33,4 +33,3 @@ jobs:
         run: |
           python -m venv /opt/venv && export PATH="/opt/venv/bin:$PATH"
           pytest tests/others/test_dependencies.py
-      

.github/workflows/pr_test_peft_backend.yml

@@ -115,14 +115,14 @@ jobs:
           -s -v \
           --make-reports=tests_models_lora_${{ matrix.config.report }} \
           tests/models/ -k "lora"
-        
-    
+
+
     - name: Failure short reports
       if: ${{ failure() }}
       run: |
         cat reports/tests_${{ matrix.config.report }}_failures_short.txt
         cat reports/tests_models_lora_${{ matrix.config.report }}_failures_short.txt
-    
+
     - name: Test suite reports artifacts
       if: ${{ always() }}
       uses: actions/upload-artifact@v2

.github/workflows/push_tests.yml

@@ -330,6 +330,7 @@ jobs:
     - name: Run example tests on GPU
       env:
         HF_TOKEN: ${{ secrets.HF_TOKEN }}
+        RUN_COMPILE: yes
       run: |
         python -m pytest -n 1 --max-worker-restart=0 --dist=loadfile -s -v -k "compile" --make-reports=tests_torch_compile_cuda tests/
     - name: Failure short reports

.github/workflows/pypi_publish.yaml

@@ -29,7 +29,7 @@ jobs:
           LATEST_BRANCH=$(python utils/fetch_latest_release_branch.py)
           echo "Latest branch: $LATEST_BRANCH"
           echo "latest_branch=$LATEST_BRANCH" >> $GITHUB_ENV
-          
+
       - name: Set latest branch output
         id: set_latest_branch
         run: echo "::set-output name=latest_branch::${{ env.latest_branch }}"
@@ -43,27 +43,27 @@ jobs:
         uses: actions/checkout@v3
         with:
           ref: ${{ needs.find-and-checkout-latest-branch.outputs.latest_branch }}
-          
+
       - name: Setup Python
         uses: actions/setup-python@v4
         with:
           python-version: "3.8"
-      
+
       - name: Install dependencies
         run: |
           python -m pip install --upgrade pip
           pip install -U setuptools wheel twine
           pip install -U torch --index-url https://download.pytorch.org/whl/cpu
           pip install -U transformers
-      
+
       - name: Build the dist files
         run: python setup.py bdist_wheel && python setup.py sdist
-      
+
       - name: Publish to the test PyPI
         env:
           TWINE_USERNAME: ${{ secrets.TEST_PYPI_USERNAME }}
           TWINE_PASSWORD: ${{ secrets.TEST_PYPI_PASSWORD }}
-        run: twine upload dist/* -r pypitest --repository-url=https://test.pypi.org/legacy/    
+        run: twine upload dist/* -r pypitest --repository-url=https://test.pypi.org/legacy/
 
       - name: Test installing diffusers and importing
         run: |

.github/workflows/run_tests_from_a_pr.yml

@@ -7,7 +7,7 @@ on:
         default: 'diffusers/diffusers-pytorch-cuda'
         description: 'Name of the Docker image'
         required: true
-      branch: 
+      branch:
         description: 'PR Branch to test on'
         required: true
       test:
@@ -34,19 +34,19 @@ jobs:
     steps:
       - name: Validate test files input
         id: validate_test_files
-        env: 
+        env:
           PY_TEST: ${{ github.event.inputs.test }}
         run: |
           if [[ ! "$PY_TEST" =~ ^tests/ ]]; then
             echo "Error: The input string must start with 'tests/'."
             exit 1
           fi
-          
+
           if [[ ! "$PY_TEST" =~ ^tests/(models|pipelines) ]]; then
             echo "Error: The input string must contain either 'models' or 'pipelines' after 'tests/'."
             exit 1
           fi
-          
+
           if [[ "$PY_TEST" == *";"* ]]; then
             echo "Error: The input string must not contain ';'."
             exit 1
@@ -60,14 +60,14 @@ jobs:
           repository: ${{ github.event.pull_request.head.repo.full_name }}
 
 
-      - name: Install pytest 
-        run: | 
+      - name: Install pytest
+        run: |
           python -m venv /opt/venv && export PATH="/opt/venv/bin:$PATH"
           python -m uv pip install -e [quality,test]
           python -m uv pip install peft
-      
+
       - name: Run tests
-        env: 
+        env:
             PY_TEST: ${{ github.event.inputs.test }}
         run: |
           pytest "$PY_TEST"

CONTRIBUTING.md

@@ -245,7 +245,7 @@ The official training examples are maintained by the Diffusers' core maintainers
 This is because of the same reasons put forward in [6. Contribute a community pipeline](#6-contribute-a-community-pipeline) for official pipelines vs. community pipelines: It is not feasible for the core maintainers to maintain all possible training methods for diffusion models.
 If the Diffusers core maintainers and the community consider a certain training paradigm to be too experimental or not popular enough, the corresponding training code should be put in the `research_projects` folder and maintained by the author.
 
-Both official training and research examples consist of a directory that contains one or more training scripts, a requirements.txt file, and a README.md file. In order for the user to make use of the
+Both official training and research examples consist of a directory that contains one or more training scripts, a `requirements.txt` file, and a `README.md` file. In order for the user to make use of the
 training examples, it is required to clone the repository:
 
 ```bash
@@ -255,7 +255,8 @@ git clone https://github.com/huggingface/diffusers
 as well as to install all additional dependencies required for training:
 
 ```bash
-pip install -r /examples/<your-example-folder>/requirements.txt
+cd diffusers
+pip install -r examples/<your-example-folder>/requirements.txt
 ```
 
 Therefore when adding an example, the `requirements.txt` file shall define all pip dependencies required for your training example so that once all those are installed, the user can run the example's training script. See, for example, the [DreamBooth `requirements.txt` file](https://github.com/huggingface/diffusers/blob/main/examples/dreambooth/requirements.txt).
@@ -502,4 +503,4 @@ $ git push --set-upstream origin your-branch-for-syncing
 
 ### Style guide
 
-For documentation strings, 🧨 Diffusers follows the [Google style](https://google.github.io/styleguide/pyguide.html).
+For documentation strings, 🧨 Diffusers follows the [Google style](https://google.github.io/styleguide/pyguide.html).

PHILOSOPHY.md

@@ -63,14 +63,14 @@ Let's walk through more detailed design decisions for each class.
 Pipelines are designed to be easy to use (therefore do not follow [*Simple over easy*](#simple-over-easy) 100%), are not feature complete, and should loosely be seen as examples of how to use [models](#models) and [schedulers](#schedulers) for inference.
 
 The following design principles are followed:
-- Pipelines follow the single-file policy. All pipelines can be found in individual directories under src/diffusers/pipelines. One pipeline folder corresponds to one diffusion paper/project/release. Multiple pipeline files can be gathered in one pipeline folder, as it’s done for [`src/diffusers/pipelines/stable-diffusion`](https://github.com/huggingface/diffusers/tree/main/src/diffusers/pipelines/stable_diffusion). If pipelines share similar functionality, one can make use of the [#Copied from mechanism](https://github.com/huggingface/diffusers/blob/125d783076e5bd9785beb05367a2d2566843a271/src/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion_img2img.py#L251).
+- Pipelines follow the single-file policy. All pipelines can be found in individual directories under src/diffusers/pipelines. One pipeline folder corresponds to one diffusion paper/project/release. Multiple pipeline files can be gathered in one pipeline folder, as it’s done for [`src/diffusers/pipelines/stable-diffusion`](https://github.com/huggingface/diffusers/tree/main/src/diffusers/pipelines/stable_diffusion). If pipelines share similar functionality, one can make use of the [# Copied from mechanism](https://github.com/huggingface/diffusers/blob/125d783076e5bd9785beb05367a2d2566843a271/src/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion_img2img.py#L251).
 - Pipelines all inherit from [`DiffusionPipeline`].
 - Every pipeline consists of different model and scheduler components, that are documented in the [`model_index.json` file](https://huggingface.co/runwayml/stable-diffusion-v1-5/blob/main/model_index.json), are accessible under the same name as attributes of the pipeline and can be shared between pipelines with [`DiffusionPipeline.components`](https://huggingface.co/docs/diffusers/main/en/api/diffusion_pipeline#diffusers.DiffusionPipeline.components) function.
 - Every pipeline should be loadable via the [`DiffusionPipeline.from_pretrained`](https://huggingface.co/docs/diffusers/main/en/api/diffusion_pipeline#diffusers.DiffusionPipeline.from_pretrained) function.
 - Pipelines should be used **only** for inference.
 - Pipelines should be very readable, self-explanatory, and easy to tweak.
 - Pipelines should be designed to build on top of each other and be easy to integrate into higher-level APIs.
-- Pipelines are **not** intended to be feature-complete user interfaces. For future complete user interfaces one should rather have a look at [InvokeAI](https://github.com/invoke-ai/InvokeAI), [Diffuzers](https://github.com/abhishekkrthakur/diffuzers), and [lama-cleaner](https://github.com/Sanster/lama-cleaner).
+- Pipelines are **not** intended to be feature-complete user interfaces. For feature-complete user interfaces one should rather have a look at [InvokeAI](https://github.com/invoke-ai/InvokeAI), [Diffuzers](https://github.com/abhishekkrthakur/diffuzers), and [lama-cleaner](https://github.com/Sanster/lama-cleaner).
 - Every pipeline should have one and only one way to run it via a `__call__` method. The naming of the `__call__` arguments should be shared across all pipelines.
 - Pipelines should be named after the task they are intended to solve.
 - In almost all cases, novel diffusion pipelines shall be implemented in a new pipeline folder/file.
@@ -81,7 +81,7 @@ Models are designed as configurable toolboxes that are natural extensions of [Py
 
 The following design principles are followed:
 - Models correspond to **a type of model architecture**. *E.g.* the [`UNet2DConditionModel`] class is used for all UNet variations that expect 2D image inputs and are conditioned on some context.
-- All models can be found in [`src/diffusers/models`](https://github.com/huggingface/diffusers/tree/main/src/diffusers/models) and every model architecture shall be defined in its file, e.g. [`unet_2d_condition.py`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/unet_2d_condition.py), [`transformer_2d.py`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/transformer_2d.py), etc...
+- All models can be found in [`src/diffusers/models`](https://github.com/huggingface/diffusers/tree/main/src/diffusers/models) and every model architecture shall be defined in its file, e.g. [`unets/unet_2d_condition.py`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/unets/unet_2d_condition.py), [`transformers/transformer_2d.py`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/transformers/transformer_2d.py), etc...
 - Models **do not** follow the single-file policy and should make use of smaller model building blocks, such as [`attention.py`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/attention.py), [`resnet.py`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/resnet.py), [`embeddings.py`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/embeddings.py), etc... **Note**: This is in stark contrast to Transformers' modeling files and shows that models do not really follow the single-file policy.
 - Models intend to expose complexity, just like PyTorch's `Module` class, and give clear error messages.
 - Models all inherit from `ModelMixin` and `ConfigMixin`.
@@ -90,7 +90,7 @@ The following design principles are followed:
 - To integrate new model checkpoints whose general architecture can be classified as an architecture that already exists in Diffusers, the existing model architecture shall be adapted to make it work with the new checkpoint. One should only create a new file if the model architecture is fundamentally different.
 - Models should be designed to be easily extendable to future changes. This can be achieved by limiting public function arguments, configuration arguments, and "foreseeing" future changes, *e.g.* it is usually better to add `string` "...type" arguments that can easily be extended to new future types instead of boolean `is_..._type` arguments. Only the minimum amount of changes shall be made to existing architectures to make a new model checkpoint work.
 - The model design is a difficult trade-off between keeping code readable and concise and supporting many model checkpoints. For most parts of the modeling code, classes shall be adapted for new model checkpoints, while there are some exceptions where it is preferred to add new classes to make sure the code is kept concise and
-readable long-term, such as [UNet blocks](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/unet_2d_blocks.py) and [Attention processors](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/attention_processor.py).
+readable long-term, such as [UNet blocks](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/unets/unet_2d_blocks.py) and [Attention processors](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/attention_processor.py).
 
 ### Schedulers
 
@@ -100,11 +100,11 @@ The following design principles are followed:
 - All schedulers are found in [`src/diffusers/schedulers`](https://github.com/huggingface/diffusers/tree/main/src/diffusers/schedulers).
 - Schedulers are **not** allowed to import from large utils files and shall be kept very self-contained.
 - One scheduler Python file corresponds to one scheduler algorithm (as might be defined in a paper).
-- If schedulers share similar functionalities, we can make use of the `#Copied from` mechanism.
+- If schedulers share similar functionalities, we can make use of the `# Copied from` mechanism.
 - Schedulers all inherit from `SchedulerMixin` and `ConfigMixin`.
 - Schedulers can be easily swapped out with the [`ConfigMixin.from_config`](https://huggingface.co/docs/diffusers/main/en/api/configuration#diffusers.ConfigMixin.from_config) method as explained in detail [here](./docs/source/en/using-diffusers/schedulers.md).
 - Every scheduler has to have a `set_num_inference_steps`, and a `step` function. `set_num_inference_steps(...)` has to be called before every denoising process, *i.e.* before `step(...)` is called.
 - Every scheduler exposes the timesteps to be "looped over" via a `timesteps` attribute, which is an array of timesteps the model will be called upon.
 - The `step(...)` function takes a predicted model output and the "current" sample (x_t) and returns the "previous", slightly more denoised sample (x_t-1).
 - Given the complexity of diffusion schedulers, the `step` function does not expose all the complexity and can be a bit of a "black box".
-- In almost all cases, novel schedulers shall be implemented in a new scheduling file.
+- In almost all cases, novel schedulers shall be implemented in a new scheduling file.

README.md

@@ -20,21 +20,11 @@ limitations under the License.
     <br>
 <p>
 <p align="center">
-    <a href="https://github.com/huggingface/diffusers/blob/main/LICENSE">
-        <img alt="GitHub" src="https://img.shields.io/github/license/huggingface/datasets.svg?color=blue">
-    </a>
-    <a href="https://github.com/huggingface/diffusers/releases">
-        <img alt="GitHub release" src="https://img.shields.io/github/release/huggingface/diffusers.svg">
-    </a>
-    <a href="https://pepy.tech/project/diffusers">
-        <img alt="GitHub release" src="https://static.pepy.tech/badge/diffusers/month">
-    </a>
-    <a href="CODE_OF_CONDUCT.md">
-        <img alt="Contributor Covenant" src="https://img.shields.io/badge/Contributor%20Covenant-2.1-4baaaa.svg">
-    </a>
-    <a href="https://twitter.com/diffuserslib">
-        <img alt="X account" src="https://img.shields.io/twitter/url/https/twitter.com/diffuserslib.svg?style=social&label=Follow%20%40diffuserslib">
-    </a>
+    <a href="https://github.com/huggingface/diffusers/blob/main/LICENSE"><img alt="GitHub" src="https://img.shields.io/github/license/huggingface/datasets.svg?color=blue"></a>
+    <a href="https://github.com/huggingface/diffusers/releases"><img alt="GitHub release" src="https://img.shields.io/github/release/huggingface/diffusers.svg"></a>
+    <a href="https://pepy.tech/project/diffusers"><img alt="GitHub release" src="https://static.pepy.tech/badge/diffusers/month"></a>
+    <a href="CODE_OF_CONDUCT.md"><img alt="Contributor Covenant" src="https://img.shields.io/badge/Contributor%20Covenant-2.1-4baaaa.svg"></a>
+    <a href="https://twitter.com/diffuserslib"><img alt="X account" src="https://img.shields.io/twitter/url/https/twitter.com/diffuserslib.svg?style=social&label=Follow%20%40diffuserslib"></a>
 </p>
 
 🤗 Diffusers is the go-to library for state-of-the-art pretrained diffusion models for generating images, audio, and even 3D structures of molecules. Whether you're looking for a simple inference solution or training your own diffusion models, 🤗 Diffusers is a modular toolbox that supports both. Our library is designed with a focus on [usability over performance](https://huggingface.co/docs/diffusers/conceptual/philosophy#usability-over-performance), [simple over easy](https://huggingface.co/docs/diffusers/conceptual/philosophy#simple-over-easy), and [customizability over abstractions](https://huggingface.co/docs/diffusers/conceptual/philosophy#tweakable-contributorfriendly-over-abstraction).
@@ -77,7 +67,7 @@ Please refer to the [How to use Stable Diffusion in Apple Silicon](https://huggi
 
 ## Quickstart
 
-Generating outputs is super easy with 🤗 Diffusers. To generate an image from text, use the `from_pretrained` method to load any pretrained diffusion model (browse the [Hub](https://huggingface.co/models?library=diffusers&sort=downloads) for 25.000+ checkpoints):
+Generating outputs is super easy with 🤗 Diffusers. To generate an image from text, use the `from_pretrained` method to load any pretrained diffusion model (browse the [Hub](https://huggingface.co/models?library=diffusers&sort=downloads) for 27.000+ checkpoints):
 
 ```python
 from diffusers import DiffusionPipeline
@@ -219,7 +209,7 @@ Also, say 👋 in our public Discord channel <a href="https://discord.gg/G7tWnz9
 - https://github.com/deep-floyd/IF
 - https://github.com/bentoml/BentoML
 - https://github.com/bmaltais/kohya_ss
-- +11.000 other amazing GitHub repositories 💪
+- +12.000 other amazing GitHub repositories 💪
 
 Thank you for using us ❤️.
 

benchmarks/run_all.py

@@ -40,7 +40,7 @@ def main():
         print(f"****** Running file: {file} ******")
 
         # Run with canonical settings.
-        if file != "benchmark_text_to_image.py":
+        if file != "benchmark_text_to_image.py" and file != "benchmark_ip_adapters.py":
             command = f"python {file}"
             run_command(command.split())
 
@@ -49,6 +49,10 @@ def main():
 
     # Run variants.
     for file in python_files:
+        # See: https://github.com/pytorch/pytorch/issues/129637
+        if file == "benchmark_ip_adapters.py":
+            continue
+
         if file == "benchmark_text_to_image.py":
             for ckpt in ALL_T2I_CKPTS:
                 command = f"python {file} --ckpt {ckpt}"

docker/diffusers-doc-builder/Dockerfile

@@ -42,7 +42,7 @@ RUN python3.10 -m pip install --no-cache-dir --upgrade pip uv==0.1.11 && \
         huggingface-hub \
         Jinja2 \
         librosa \
-        numpy \
+        numpy==1.26.4 \
         scipy \
         tensorboard \
         transformers \

docker/diffusers-flax-cpu/Dockerfile

@@ -40,7 +40,7 @@ RUN python3 -m pip install --no-cache-dir --upgrade pip uv==0.1.11 && \
         huggingface-hub \
         Jinja2 \
         librosa \
-        numpy \
+        numpy==1.26.4 \
         scipy \
         tensorboard \
         transformers

docker/diffusers-flax-tpu/Dockerfile

@@ -41,8 +41,8 @@ RUN python3 -m pip install --no-cache-dir --upgrade pip uv==0.1.11 && \
         hf-doc-builder \
         huggingface-hub \
         Jinja2 \
-        librosa \        
-        numpy \
+        librosa \
+        numpy==1.26.4 \
         scipy \
         tensorboard \
         transformers

docker/diffusers-onnxruntime-cpu/Dockerfile

@@ -40,7 +40,7 @@ RUN python3 -m pip install --no-cache-dir --upgrade pip uv==0.1.11 && \
         huggingface-hub \
         Jinja2 \
         librosa \
-        numpy \
+        numpy==1.26.4 \
         scipy \
         tensorboard \
         transformers

docker/diffusers-onnxruntime-cuda/Dockerfile

@@ -40,7 +40,7 @@ RUN python3.10 -m pip install --no-cache-dir --upgrade pip uv==0.1.11 && \
         huggingface-hub \
         Jinja2 \
         librosa \
-        numpy \
+        numpy==1.26.4 \
         scipy \
         tensorboard \
         transformers

docker/diffusers-pytorch-compile-cuda/Dockerfile

@@ -16,30 +16,31 @@ RUN apt install -y bash \
     ca-certificates \
     libsndfile1-dev \
     libgl1 \
-    python3.10 \
+    python3.9 \
+    python3.9-dev \
     python3-pip \
-    python3.10-venv && \
+    python3.9-venv && \
     rm -rf /var/lib/apt/lists
 
 # make sure to use venv
-RUN python3.10 -m venv /opt/venv
+RUN python3.9 -m venv /opt/venv
 ENV PATH="/opt/venv/bin:$PATH"
 
 # pre-install the heavy dependencies (these can later be overridden by the deps from setup.py)
-RUN python3.10 -m pip install --no-cache-dir --upgrade pip uv==0.1.11 && \
-    python3.10 -m uv pip install --no-cache-dir \
+RUN python3.9 -m pip install --no-cache-dir --upgrade pip uv==0.1.11 && \
+    python3.9 -m uv pip install --no-cache-dir \
     torch \
     torchvision \
     torchaudio \
     invisible_watermark && \
-    python3.10 -m pip install --no-cache-dir \
+    python3.9 -m pip install --no-cache-dir \
     accelerate \
     datasets \
     hf-doc-builder \
     huggingface-hub \
     Jinja2 \
     librosa \
-    numpy \
+    numpy==1.26.4 \
     scipy \
     tensorboard \
     transformers

docker/diffusers-pytorch-cpu/Dockerfile

@@ -40,7 +40,7 @@ RUN python3.10 -m pip install --no-cache-dir --upgrade pip uv==0.1.11 && \
         huggingface-hub \
         Jinja2 \
         librosa \
-        numpy \
+        numpy==1.26.4 \
         scipy \
         tensorboard \
         transformers matplotlib

docker/diffusers-pytorch-cuda/Dockerfile

@@ -39,7 +39,7 @@ RUN python3.10 -m pip install --no-cache-dir --upgrade pip uv==0.1.11 && \
     huggingface-hub \
     Jinja2 \
     librosa \
-    numpy \
+    numpy==1.26.4 \
     scipy \
     tensorboard \
     transformers \

docker/diffusers-pytorch-xformers-cuda/Dockerfile

@@ -39,7 +39,7 @@ RUN python3.10 -m pip install --no-cache-dir --upgrade pip uv==0.1.11 && \
         huggingface-hub \
         Jinja2 \
         librosa \
-        numpy \
+        numpy==1.26.4 \
         scipy \
         tensorboard \
         transformers \

docs/source/en/_toctree.yml

@@ -21,6 +21,8 @@
     title: Load LoRAs for inference
   - local: tutorials/fast_diffusion
     title: Accelerate inference of text-to-image diffusion models
+  - local: tutorials/inference_with_big_models
+    title: Working with big models
   title: Tutorials
 - sections:
   - local: using-diffusers/loading
@@ -81,6 +83,8 @@
     title: Kandinsky
   - local: using-diffusers/ip_adapter
     title: IP-Adapter
+  - local: using-diffusers/pag
+    title: PAG
   - local: using-diffusers/controlnet
     title: ControlNet
   - local: using-diffusers/t2i_adapter
@@ -253,6 +257,10 @@
       title: PriorTransformer
     - local: api/models/controlnet
       title: ControlNetModel
+    - local: api/models/controlnet_hunyuandit
+      title: HunyuanDiT2DControlNetModel
+    - local: api/models/controlnet_sd3
+      title: SD3ControlNetModel
     title: Models
   - isExpanded: false
     sections:
@@ -276,6 +284,10 @@
       title: Consistency Models
     - local: api/pipelines/controlnet
       title: ControlNet
+    - local: api/pipelines/controlnet_hunyuandit
+      title: ControlNet with Hunyuan-DiT
+    - local: api/pipelines/controlnet_sd3
+      title: ControlNet with Stable Diffusion 3
     - local: api/pipelines/controlnet_sdxl
       title: ControlNet with Stable Diffusion XL
     - local: api/pipelines/controlnetxs
@@ -318,6 +330,8 @@
       title: MultiDiffusion
     - local: api/pipelines/musicldm
       title: MusicLDM
+    - local: api/pipelines/pag
+      title: PAG
     - local: api/pipelines/paint_by_example
       title: Paint by Example
     - local: api/pipelines/pia

docs/source/en/api/attnprocessor.md

@@ -41,12 +41,6 @@ An attention processor is a class for applying different types of attention mech
 ## FusedAttnProcessor2_0
 [[autodoc]] models.attention_processor.FusedAttnProcessor2_0
 
-## LoRAAttnAddedKVProcessor
-[[autodoc]] models.attention_processor.LoRAAttnAddedKVProcessor
-
-## LoRAXFormersAttnProcessor
-[[autodoc]] models.attention_processor.LoRAXFormersAttnProcessor
-
 ## SlicedAttnProcessor
 [[autodoc]] models.attention_processor.SlicedAttnProcessor
 

docs/source/en/api/loaders/single_file.md

@@ -35,6 +35,7 @@ The [`~loaders.FromSingleFileMixin.from_single_file`] method allows you to load:
 - [`StableDiffusionXLInstructPix2PixPipeline`]
 - [`StableDiffusionXLControlNetPipeline`]
 - [`StableDiffusionXLKDiffusionPipeline`]
+- [`StableDiffusion3Pipeline`]
 - [`LatentConsistencyModelPipeline`]
 - [`LatentConsistencyModelImg2ImgPipeline`]
 - [`StableDiffusionControlNetXSPipeline`]
@@ -49,6 +50,7 @@ The [`~loaders.FromSingleFileMixin.from_single_file`] method allows you to load:
 - [`StableCascadeUNet`]
 - [`AutoencoderKL`]
 - [`ControlNetModel`]
+- [`SD3Transformer2DModel`]
 
 ## FromSingleFileMixin
 

docs/source/en/api/models/autoencoderkl.md

@@ -21,7 +21,7 @@ The abstract from the paper is:
 ## Loading from the original format
 
 By default the [`AutoencoderKL`] should be loaded with [`~ModelMixin.from_pretrained`], but it can also be loaded
-from the original format using [`FromOriginalVAEMixin.from_single_file`] as follows:
+from the original format using [`FromOriginalModelMixin.from_single_file`] as follows:
 
 ```py
 from diffusers import AutoencoderKL

docs/source/en/api/models/controlnet.md

@@ -21,7 +21,7 @@ The abstract from the paper is:
 ## Loading from the original format
 
 By default the [`ControlNetModel`] should be loaded with [`~ModelMixin.from_pretrained`], but it can also be loaded
-from the original format using [`FromOriginalControlnetMixin.from_single_file`] as follows:
+from the original format using [`FromOriginalModelMixin.from_single_file`] as follows:
 
 ```py
 from diffusers import StableDiffusionControlNetPipeline, ControlNetModel

docs/source/en/api/models/controlnet_hunyuandit.md

@@ -0,0 +1,37 @@
+<!--Copyright 2024 The HuggingFace Team and Tencent Hunyuan Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# HunyuanDiT2DControlNetModel
+
+HunyuanDiT2DControlNetModel is an implementation of ControlNet for [Hunyuan-DiT](https://arxiv.org/abs/2405.08748).
+
+ControlNet was introduced in [Adding Conditional Control to Text-to-Image Diffusion Models](https://huggingface.co/papers/2302.05543) by Lvmin Zhang, Anyi Rao, and Maneesh Agrawala.
+
+With a ControlNet model, you can provide an additional control image to condition and control Hunyuan-DiT generation. For example, if you provide a depth map, the ControlNet model generates an image that'll preserve the spatial information from the depth map. It is a more flexible and accurate way to control the image generation process.
+
+The abstract from the paper is:
+
+*We present ControlNet, a neural network architecture to add spatial conditioning controls to large, pretrained text-to-image diffusion models. ControlNet locks the production-ready large diffusion models, and reuses their deep and robust encoding layers pretrained with billions of images as a strong backbone to learn a diverse set of conditional controls. The neural architecture is connected with "zero convolutions" (zero-initialized convolution layers) that progressively grow the parameters from zero and ensure that no harmful noise could affect the finetuning. We test various conditioning controls, eg, edges, depth, segmentation, human pose, etc, with Stable Diffusion, using single or multiple conditions, with or without prompts. We show that the training of ControlNets is robust with small (<50k) and large (>1m) datasets. Extensive results show that ControlNet may facilitate wider applications to control image diffusion models.*
+
+This code is implemented by Tencent Hunyuan Team. You can find pre-trained checkpoints for Hunyuan-DiT ControlNets on [Tencent Hunyuan](https://huggingface.co/Tencent-Hunyuan).
+
+## Example For Loading HunyuanDiT2DControlNetModel
+
+```py
+from diffusers import HunyuanDiT2DControlNetModel
+import torch
+controlnet = HunyuanDiT2DControlNetModel.from_pretrained("Tencent-Hunyuan/HunyuanDiT-v1.1-ControlNet-Diffusers-Pose", torch_dtype=torch.float16)
+```
+
+## HunyuanDiT2DControlNetModel
+
+[[autodoc]] HunyuanDiT2DControlNetModel

docs/source/en/api/models/controlnet_sd3.md

@@ -0,0 +1,42 @@
+<!--Copyright 2024 The HuggingFace Team and The InstantX Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# SD3ControlNetModel
+
+SD3ControlNetModel is an implementation of ControlNet for Stable Diffusion 3.
+
+The ControlNet model was introduced in [Adding Conditional Control to Text-to-Image Diffusion Models](https://huggingface.co/papers/2302.05543) by Lvmin Zhang, Anyi Rao, Maneesh Agrawala. It provides a greater degree of control over text-to-image generation by conditioning the model on additional inputs such as edge maps, depth maps, segmentation maps, and keypoints for pose detection.
+
+The abstract from the paper is:
+
+*We present ControlNet, a neural network architecture to add spatial conditioning controls to large, pretrained text-to-image diffusion models. ControlNet locks the production-ready large diffusion models, and reuses their deep and robust encoding layers pretrained with billions of images as a strong backbone to learn a diverse set of conditional controls. The neural architecture is connected with "zero convolutions" (zero-initialized convolution layers) that progressively grow the parameters from zero and ensure that no harmful noise could affect the finetuning. We test various conditioning controls, eg, edges, depth, segmentation, human pose, etc, with Stable Diffusion, using single or multiple conditions, with or without prompts. We show that the training of ControlNets is robust with small (<50k) and large (>1m) datasets. Extensive results show that ControlNet may facilitate wider applications to control image diffusion models.*
+
+## Loading from the original format
+
+By default the [`SD3ControlNetModel`] should be loaded with [`~ModelMixin.from_pretrained`].
+
+```py
+from diffusers import StableDiffusion3ControlNetPipeline
+from diffusers.models import SD3ControlNetModel, SD3MultiControlNetModel
+
+controlnet = SD3ControlNetModel.from_pretrained("InstantX/SD3-Controlnet-Canny")
+pipe = StableDiffusion3ControlNetPipeline.from_pretrained("stabilityai/stable-diffusion-3-medium-diffusers", controlnet=controlnet)
+```
+
+## SD3ControlNetModel
+
+[[autodoc]] SD3ControlNetModel
+
+## SD3ControlNetOutput
+
+[[autodoc]] models.controlnet_sd3.SD3ControlNetOutput
+

docs/source/en/api/models/pixart_transformer2d.md

@@ -12,7 +12,7 @@ specific language governing permissions and limitations under the License.
 
 # PixArtTransformer2DModel
 
-A Transformer model for image-like data from [PixArt-Alpha](https://huggingface.co/papers/2310.00426) and [PixArt-Sigma](https://huggingface.co/papers/2403.04692). 
+A Transformer model for image-like data from [PixArt-Alpha](https://huggingface.co/papers/2310.00426) and [PixArt-Sigma](https://huggingface.co/papers/2403.04692).
 
 ## PixArtTransformer2DModel
 

docs/source/en/api/models/sd3_transformer2d.md

@@ -12,7 +12,7 @@ specific language governing permissions and limitations under the License.
 
 # SD3 Transformer Model
 
-The Transformer model introduced in [Stable Diffusion 3](https://hf.co/papers/2403.03206). Its novelty lies in the MMDiT transformer block. 
+The Transformer model introduced in [Stable Diffusion 3](https://hf.co/papers/2403.03206). Its novelty lies in the MMDiT transformer block.
 
 ## SD3Transformer2DModel
 

docs/source/en/api/models/transformer2d.md

@@ -38,4 +38,4 @@ It is assumed one of the input classes is the masked latent pixel. The predicted
 
 ## Transformer2DModelOutput
 
-[[autodoc]] models.transformers.transformer_2d.Transformer2DModelOutput
+[[autodoc]] models.modeling_outputs.Transformer2DModelOutput

docs/source/en/api/pipelines/animatediff.md

@@ -78,7 +78,6 @@ output = pipe(
 )
 frames = output.frames[0]
 export_to_gif(frames, "animation.gif")
-
 ```
 
 Here are some sample outputs:
@@ -303,7 +302,6 @@ output = pipe(
 )
 frames = output.frames[0]
 export_to_gif(frames, "animation.gif")
-
 ```
 
 <table>
@@ -378,7 +376,6 @@ output = pipe(
 )
 frames = output.frames[0]
 export_to_gif(frames, "animation.gif")
-
 ```
 
 <table>

docs/source/en/api/pipelines/audioldm2.md

@@ -20,8 +20,8 @@ The abstract of the paper is the following:
 
 *Although audio generation shares commonalities across different types of audio, such as speech, music, and sound effects, designing models for each type requires careful consideration of specific objectives and biases that can significantly differ from those of other types. To bring us closer to a unified perspective of audio generation, this paper proposes a framework that utilizes the same learning method for speech, music, and sound effect generation. Our framework introduces a general representation of audio, called "language of audio" (LOA). Any audio can be translated into LOA based on AudioMAE, a self-supervised pre-trained representation learning model. In the generation process, we translate any modalities into LOA by using a GPT-2 model, and we perform self-supervised audio generation learning with a latent diffusion model conditioned on LOA. The proposed framework naturally brings advantages such as in-context learning abilities and reusable self-supervised pretrained AudioMAE and latent diffusion models. Experiments on the major benchmarks of text-to-audio, text-to-music, and text-to-speech demonstrate state-of-the-art or competitive performance against previous approaches. Our code, pretrained model, and demo are available at [this https URL](https://audioldm.github.io/audioldm2).*
 
-This pipeline was contributed by [sanchit-gandhi](https://huggingface.co/sanchit-gandhi) and [Nguyễn Công Tú Anh](https://github.com/tuanh123789). The original codebase can be 
-found at [haoheliu/audioldm2](https://github.com/haoheliu/audioldm2). 
+This pipeline was contributed by [sanchit-gandhi](https://huggingface.co/sanchit-gandhi) and [Nguyễn Công Tú Anh](https://github.com/tuanh123789). The original codebase can be
+found at [haoheliu/audioldm2](https://github.com/haoheliu/audioldm2).
 
 ## Tips
 

docs/source/en/api/pipelines/blip_diffusion.md

@@ -12,7 +12,7 @@ specific language governing permissions and limitations under the License.
 
 # BLIP-Diffusion
 
-BLIP-Diffusion was proposed in [BLIP-Diffusion: Pre-trained Subject Representation for Controllable Text-to-Image Generation and Editing](https://arxiv.org/abs/2305.14720). It enables zero-shot subject-driven generation and control-guided zero-shot generation. 
+BLIP-Diffusion was proposed in [BLIP-Diffusion: Pre-trained Subject Representation for Controllable Text-to-Image Generation and Editing](https://arxiv.org/abs/2305.14720). It enables zero-shot subject-driven generation and control-guided zero-shot generation.
 
 
 The abstract from the paper is:

docs/source/en/api/pipelines/controlnet_hunyuandit.md

@@ -0,0 +1,36 @@
+<!--Copyright 2024 The HuggingFace Team and Tencent Hunyuan Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# ControlNet with Hunyuan-DiT
+
+HunyuanDiTControlNetPipeline is an implementation of ControlNet for [Hunyuan-DiT](https://arxiv.org/abs/2405.08748).
+
+ControlNet was introduced in [Adding Conditional Control to Text-to-Image Diffusion Models](https://huggingface.co/papers/2302.05543) by Lvmin Zhang, Anyi Rao, and Maneesh Agrawala.
+
+With a ControlNet model, you can provide an additional control image to condition and control Hunyuan-DiT generation. For example, if you provide a depth map, the ControlNet model generates an image that'll preserve the spatial information from the depth map. It is a more flexible and accurate way to control the image generation process.
+
+The abstract from the paper is:
+
+*We present ControlNet, a neural network architecture to add spatial conditioning controls to large, pretrained text-to-image diffusion models. ControlNet locks the production-ready large diffusion models, and reuses their deep and robust encoding layers pretrained with billions of images as a strong backbone to learn a diverse set of conditional controls. The neural architecture is connected with "zero convolutions" (zero-initialized convolution layers) that progressively grow the parameters from zero and ensure that no harmful noise could affect the finetuning. We test various conditioning controls, eg, edges, depth, segmentation, human pose, etc, with Stable Diffusion, using single or multiple conditions, with or without prompts. We show that the training of ControlNets is robust with small (<50k) and large (>1m) datasets. Extensive results show that ControlNet may facilitate wider applications to control image diffusion models.*
+
+This code is implemented by Tencent Hunyuan Team. You can find pre-trained checkpoints for Hunyuan-DiT ControlNets on [Tencent Hunyuan](https://huggingface.co/Tencent-Hunyuan).
+
+<Tip>
+
+Make sure to check out the Schedulers [guide](../../using-diffusers/schedulers) to learn how to explore the tradeoff between scheduler speed and quality, and see the [reuse components across pipelines](../../using-diffusers/loading#reuse-components-across-pipelines) section to learn how to efficiently load the same components into multiple pipelines.
+
+</Tip>
+
+## HunyuanDiTControlNetPipeline
+[[autodoc]] HunyuanDiTControlNetPipeline
+	- all
+	- __call__

docs/source/en/api/pipelines/controlnet_sd3.md

@@ -0,0 +1,39 @@
+<!--Copyright 2023 The HuggingFace Team and The InstantX Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# ControlNet with Stable Diffusion 3
+
+StableDiffusion3ControlNetPipeline is an implementation of ControlNet for Stable Diffusion 3.
+
+ControlNet was introduced in [Adding Conditional Control to Text-to-Image Diffusion Models](https://huggingface.co/papers/2302.05543) by Lvmin Zhang, Anyi Rao, and Maneesh Agrawala.
+
+With a ControlNet model, you can provide an additional control image to condition and control Stable Diffusion generation. For example, if you provide a depth map, the ControlNet model generates an image that'll preserve the spatial information from the depth map. It is a more flexible and accurate way to control the image generation process.
+
+The abstract from the paper is:
+
+*We present ControlNet, a neural network architecture to add spatial conditioning controls to large, pretrained text-to-image diffusion models. ControlNet locks the production-ready large diffusion models, and reuses their deep and robust encoding layers pretrained with billions of images as a strong backbone to learn a diverse set of conditional controls. The neural architecture is connected with "zero convolutions" (zero-initialized convolution layers) that progressively grow the parameters from zero and ensure that no harmful noise could affect the finetuning. We test various conditioning controls, eg, edges, depth, segmentation, human pose, etc, with Stable Diffusion, using single or multiple conditions, with or without prompts. We show that the training of ControlNets is robust with small (<50k) and large (>1m) datasets. Extensive results show that ControlNet may facilitate wider applications to control image diffusion models.*
+
+This code is implemented by [The InstantX Team](https://huggingface.co/InstantX). You can find pre-trained checkpoints for SD3-ControlNet on [The InstantX Team](https://huggingface.co/InstantX) Hub profile.
+
+<Tip>
+
+Make sure to check out the Schedulers [guide](../../using-diffusers/schedulers) to learn how to explore the tradeoff between scheduler speed and quality, and see the [reuse components across pipelines](../../using-diffusers/loading#reuse-components-across-pipelines) section to learn how to efficiently load the same components into multiple pipelines.
+
+</Tip>
+
+## StableDiffusion3ControlNetPipeline
+[[autodoc]] StableDiffusion3ControlNetPipeline
+	- all
+	- __call__
+
+## StableDiffusion3PipelineOutput
+[[autodoc]] pipelines.stable_diffusion_3.pipeline_output.StableDiffusion3PipelineOutput

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

@@ -1,4 +1,4 @@
-<!--Copyright 2024 The HuggingFace Team. All rights reserved.
+<!--Copyright 2024 The HuggingFace Team and Tencent Hunyuan Team. All rights reserved.
 
 Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
 the License. You may obtain a copy of the License at
@@ -34,9 +34,15 @@ Make sure to check out the Schedulers [guide](../../using-diffusers/schedulers.m
 
 </Tip>
 
+<Tip>
+
+You can further improve generation quality by passing the generated image from [`HungyuanDiTPipeline`] to the [SDXL refiner](../../using-diffusers/sdxl#base-to-refiner-model) model.
+
+</Tip>
+
 ## Optimization
 
-You can optimize the pipeline's runtime and memory consumption with torch.compile and feed-forward chunking. To learn about other optimization methods, check out the [Speed up inference](../../optimization/fp16) and [Reduce memory usage](../../optimization/memory) guides. 
+You can optimize the pipeline's runtime and memory consumption with torch.compile and feed-forward chunking. To learn about other optimization methods, check out the [Speed up inference](../../optimization/fp16) and [Reduce memory usage](../../optimization/memory) guides.
 
 ### Inference
 
@@ -46,7 +52,7 @@ First, load the pipeline:
 
 ```python
 from diffusers import HunyuanDiTPipeline
-import torch 
+import torch
 
 pipeline = HunyuanDiTPipeline.from_pretrained(
 	"Tencent-Hunyuan/HunyuanDiT-Diffusers", torch_dtype=torch.float16
@@ -78,7 +84,7 @@ Without torch.compile(): Average inference time: 20.570 seconds.
 
 ### Memory optimization
 
-By loading the T5 text encoder in 8 bits, you can run the pipeline in just under 6 GBs of GPU VRAM. Refer to [this script](https://gist.github.com/sayakpaul/3154605f6af05b98a41081aaba5ca43e) for details. 
+By loading the T5 text encoder in 8 bits, you can run the pipeline in just under 6 GBs of GPU VRAM. Refer to [this script](https://gist.github.com/sayakpaul/3154605f6af05b98a41081aaba5ca43e) for details.
 
 Furthermore, you can use the [`~HunyuanDiT2DModel.enable_forward_chunking`] method to reduce memory usage. Feed-forward chunking runs the feed-forward layers in a transformer block in a loop instead of all at once. This gives you a trade-off between memory consumption and inference runtime.
 
@@ -92,4 +98,4 @@ Furthermore, you can use the [`~HunyuanDiT2DModel.enable_forward_chunking`] meth
 [[autodoc]] HunyuanDiTPipeline
 	- all
 	- __call__
-	
+

docs/source/en/api/pipelines/kandinsky3.md

@@ -11,7 +11,7 @@ specific language governing permissions and limitations under the License.
 
 Kandinsky 3 is created by [Vladimir Arkhipkin](https://github.com/oriBetelgeuse),[Anastasia Maltseva](https://github.com/NastyaMittseva),[Igor Pavlov](https://github.com/boomb0om),[Andrei Filatov](https://github.com/anvilarth),[Arseniy Shakhmatov](https://github.com/cene555),[Andrey Kuznetsov](https://github.com/kuznetsoffandrey),[Denis Dimitrov](https://github.com/denndimitrov), [Zein Shaheen](https://github.com/zeinsh)
 
-The description from it's Github page:
+The description from it's GitHub page:
 
 *Kandinsky 3.0 is an open-source text-to-image diffusion model built upon the Kandinsky2-x model family. In comparison to its predecessors, enhancements have been made to the text understanding and visual quality of the model, achieved by increasing the size of the text encoder and Diffusion U-Net models, respectively.*
 

docs/source/en/api/pipelines/pag.md

@@ -0,0 +1,46 @@
+<!--Copyright 2024 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# Perturbed-Attention Guidance
+
+[Perturbed-Attention Guidance (PAG)](https://ku-cvlab.github.io/Perturbed-Attention-Guidance/) is a new diffusion sampling guidance that improves sample quality across both unconditional and conditional settings, achieving this without requiring further training or the integration of external modules.
+
+PAG was introduced in [Self-Rectifying Diffusion Sampling with Perturbed-Attention Guidance](https://huggingface.co/papers/2403.17377) by Donghoon Ahn, Hyoungwon Cho, Jaewon Min, Wooseok Jang, Jungwoo Kim, SeonHwa Kim, Hyun Hee Park, Kyong Hwan Jin and Seungryong Kim.
+
+The abstract from the paper is:
+
+*Recent studies have demonstrated that diffusion models are capable of generating high-quality samples, but their quality heavily depends on sampling guidance techniques, such as classifier guidance (CG) and classifier-free guidance (CFG). These techniques are often not applicable in unconditional generation or in various downstream tasks such as image restoration. In this paper, we propose a novel sampling guidance, called Perturbed-Attention Guidance (PAG), which improves diffusion sample quality across both unconditional and conditional settings, achieving this without requiring additional training or the integration of external modules. PAG is designed to progressively enhance the structure of samples throughout the denoising process. It involves generating intermediate samples with degraded structure by substituting selected self-attention maps in diffusion U-Net with an identity matrix, by considering the self-attention mechanisms' ability to capture structural information, and guiding the denoising process away from these degraded samples. In both ADM and Stable Diffusion, PAG surprisingly improves sample quality in conditional and even unconditional scenarios. Moreover, PAG significantly improves the baseline performance in various downstream tasks where existing guidances such as CG or CFG cannot be fully utilized, including ControlNet with empty prompts and image restoration such as inpainting and deblurring.*
+
+## StableDiffusionPAGPipeline
+[[autodoc]] StableDiffusionPAGPipeline
+	- all
+	- __call__
+
+## StableDiffusionXLPAGPipeline
+[[autodoc]] StableDiffusionXLPAGPipeline
+	- all
+	- __call__
+
+## StableDiffusionXLPAGImg2ImgPipeline
+[[autodoc]] StableDiffusionXLPAGImg2ImgPipeline
+	- all
+	- __call__
+
+## StableDiffusionXLPAGInpaintPipeline
+[[autodoc]] StableDiffusionXLPAGInpaintPipeline
+	- all
+	- __call__
+
+## StableDiffusionXLControlNetPAGPipeline
+[[autodoc]] StableDiffusionXLControlNetPAGPipeline
+	- all
+	- __call__

docs/source/en/api/pipelines/pixart_sigma.md

@@ -37,6 +37,12 @@ Make sure to check out the Schedulers [guide](../../using-diffusers/schedulers)
 
 </Tip>
 
+<Tip>
+
+You can further improve generation quality by passing the generated image from [`PixArtSigmaPipeline`] to the [SDXL refiner](../../using-diffusers/sdxl#base-to-refiner-model) model.
+
+</Tip>
+
 ## Inference with under 8GB GPU VRAM
 
 Run the [`PixArtSigmaPipeline`] with under 8GB GPU VRAM by loading the text encoder in 8-bit precision. Let's walk through a full-fledged example.

docs/source/en/api/pipelines/stable_diffusion/k_diffusion.md

@@ -12,7 +12,7 @@ specific language governing permissions and limitations under the License.
 
 # K-Diffusion
 
-[k-diffusion](https://github.com/crowsonkb/k-diffusion) is a popular library created by [Katherine Crowson](https://github.com/crowsonkb/). We provide `StableDiffusionKDiffusionPipeline` and `StableDiffusionXLKDiffusionPipeline` that allow you to run Stable DIffusion with samplers from k-diffusion. 
+[k-diffusion](https://github.com/crowsonkb/k-diffusion) is a popular library created by [Katherine Crowson](https://github.com/crowsonkb/). We provide `StableDiffusionKDiffusionPipeline` and `StableDiffusionXLKDiffusionPipeline` that allow you to run Stable DIffusion with samplers from k-diffusion.
 
 Note that most the samplers from k-diffusion are implemented in Diffusers and we recommend using existing schedulers. You can find a mapping between k-diffusion samplers and schedulers in Diffusers [here](https://huggingface.co/docs/diffusers/api/schedulers/overview)
 

docs/source/en/api/pipelines/stable_diffusion/ldm3d_diffusion.md

@@ -12,11 +12,11 @@ specific language governing permissions and limitations under the License.
 
 # Text-to-(RGB, depth)
 
-LDM3D was proposed in [LDM3D: Latent Diffusion Model for 3D](https://huggingface.co/papers/2305.10853) by Gabriela Ben Melech Stan, Diana Wofk, Scottie Fox, Alex Redden, Will Saxton, Jean Yu, Estelle Aflalo, Shao-Yen Tseng, Fabio Nonato, Matthias Muller, and Vasudev Lal. LDM3D generates an image and a depth map from a given text prompt unlike the existing text-to-image diffusion models such as [Stable Diffusion](./overview) which only generates an image. With almost the same number of parameters, LDM3D achieves to create a latent space that can compress both the RGB images and the depth maps. 
+LDM3D was proposed in [LDM3D: Latent Diffusion Model for 3D](https://huggingface.co/papers/2305.10853) by Gabriela Ben Melech Stan, Diana Wofk, Scottie Fox, Alex Redden, Will Saxton, Jean Yu, Estelle Aflalo, Shao-Yen Tseng, Fabio Nonato, Matthias Muller, and Vasudev Lal. LDM3D generates an image and a depth map from a given text prompt unlike the existing text-to-image diffusion models such as [Stable Diffusion](./overview) which only generates an image. With almost the same number of parameters, LDM3D achieves to create a latent space that can compress both the RGB images and the depth maps.
 
 Two checkpoints are available for use:
 - [ldm3d-original](https://huggingface.co/Intel/ldm3d). The original checkpoint used in the [paper](https://arxiv.org/pdf/2305.10853.pdf)
-- [ldm3d-4c](https://huggingface.co/Intel/ldm3d-4c). The new version of LDM3D using 4 channels inputs instead of 6-channels inputs and finetuned on higher resolution images. 
+- [ldm3d-4c](https://huggingface.co/Intel/ldm3d-4c). The new version of LDM3D using 4 channels inputs instead of 6-channels inputs and finetuned on higher resolution images.
 
 
 The abstract from the paper is:
@@ -44,7 +44,7 @@ Make sure to check out the Stable Diffusion [Tips](overview#tips) section to lea
 
 # Upscaler
 
-[LDM3D-VR](https://arxiv.org/pdf/2311.03226.pdf) is an extended version of LDM3D. 
+[LDM3D-VR](https://arxiv.org/pdf/2311.03226.pdf) is an extended version of LDM3D.
 
 The abstract from the paper is:
 *Latent diffusion models have proven to be state-of-the-art in the creation and manipulation of visual outputs. However, as far as we know, the generation of depth maps jointly with RGB is still limited. We introduce LDM3D-VR, a suite of diffusion models targeting virtual reality development that includes LDM3D-pano and LDM3D-SR. These models enable the generation of panoramic RGBD based on textual prompts and the upscaling of low-resolution inputs to high-resolution RGBD, respectively. Our models are fine-tuned from existing pretrained models on datasets containing panoramic/high-resolution RGB images, depth maps and captions. Both models are evaluated in comparison to existing related methods*

docs/source/en/api/pipelines/stable_diffusion/stable_diffusion_2.md

@@ -48,7 +48,7 @@ from diffusers import DiffusionPipeline, DPMSolverMultistepScheduler
 import torch
 
 repo_id = "stabilityai/stable-diffusion-2-base"
-pipe = DiffusionPipeline.from_pretrained(repo_id, torch_dtype=torch.float16, revision="fp16")
+pipe = DiffusionPipeline.from_pretrained(repo_id, torch_dtype=torch.float16, variant="fp16")
 
 pipe.scheduler = DPMSolverMultistepScheduler.from_config(pipe.scheduler.config)
 pipe = pipe.to("cuda")
@@ -72,7 +72,7 @@ init_image = load_image(img_url).resize((512, 512))
 mask_image = load_image(mask_url).resize((512, 512))
 
 repo_id = "stabilityai/stable-diffusion-2-inpainting"
-pipe = DiffusionPipeline.from_pretrained(repo_id, torch_dtype=torch.float16, revision="fp16")
+pipe = DiffusionPipeline.from_pretrained(repo_id, torch_dtype=torch.float16, variant="fp16")
 
 pipe.scheduler = DPMSolverMultistepScheduler.from_config(pipe.scheduler.config)
 pipe = pipe.to("cuda")

docs/source/en/api/pipelines/stable_diffusion/stable_diffusion_3.md

@@ -21,9 +21,9 @@ The abstract from the paper is:
 
 ## Usage Example
 
-_As the model is gated, before using it with diffusers you first need to go to the [Stable Diffusion 3 Medium Hugging Face page](https://huggingface.co/stabilityai/stable-diffusion-3-medium-diffusers), fill in the form and accept the gate. Once you are in, you need to login so that your system knows you’ve accepted the gate._ 
+_As the model is gated, before using it with diffusers you first need to go to the [Stable Diffusion 3 Medium Hugging Face page](https://huggingface.co/stabilityai/stable-diffusion-3-medium-diffusers), fill in the form and accept the gate. Once you are in, you need to login so that your system knows you’ve accepted the gate._
 
-Use the command below to log in: 
+Use the command below to log in:
 
 ```bash
 huggingface-cli login
@@ -35,7 +35,6 @@ The SD3 pipeline uses three text encoders to generate an image. Model offloading
 
 </Tip>
 
-
 ```python
 import torch
 from diffusers import StableDiffusion3Pipeline
@@ -197,6 +196,68 @@ image.save("sd3_hello_world.png")
 
 Check out the full script [here](https://gist.github.com/sayakpaul/508d89d7aad4f454900813da5d42ca97).
 
+## Using Long Prompts with the T5 Text Encoder
+
+By default, the T5 Text Encoder prompt uses a maximum sequence length of `256`. This can be adjusted by setting the `max_sequence_length` to accept fewer or more tokens. Keep in mind that longer sequences require additional resources and result in longer generation times, such as during batch inference.
+
+```python
+prompt = "A whimsical and creative image depicting a hybrid creature that is a mix of a waffle and a hippopotamus, basking in a river of melted butter amidst a breakfast-themed landscape. It features the distinctive, bulky body shape of a hippo. However, instead of the usual grey skin, the creature’s body resembles a golden-brown, crispy waffle fresh off the griddle. The skin is textured with the familiar grid pattern of a waffle, each square filled with a glistening sheen of syrup. The environment combines the natural habitat of a hippo with elements of a breakfast table setting, a river of warm, melted butter, with oversized utensils or plates peeking out from the lush, pancake-like foliage in the background, a towering pepper mill standing in for a tree.  As the sun rises in this fantastical world, it casts a warm, buttery glow over the scene. The creature, content in its butter river, lets out a yawn. Nearby, a flock of birds take flight"
+
+image = pipe(
+    prompt=prompt,
+    negative_prompt="",
+    num_inference_steps=28,
+    guidance_scale=4.5,
+    max_sequence_length=512,
+).images[0]
+```
+
+### Sending a different prompt to the T5 Text Encoder
+
+You can send a different prompt to the CLIP Text Encoders and the T5 Text Encoder to prevent the prompt from being truncated by the CLIP Text Encoders and to improve generation.
+
+<Tip>
+
+The prompt with the CLIP Text Encoders is still truncated to the 77 token limit.
+
+</Tip>
+
+```python
+prompt = "A whimsical and creative image depicting a hybrid creature that is a mix of a waffle and a hippopotamus, basking in a river of melted butter amidst a breakfast-themed landscape. A river of warm, melted butter, pancake-like foliage in the background, a towering pepper mill standing in for a tree."
+
+prompt_3 = "A whimsical and creative image depicting a hybrid creature that is a mix of a waffle and a hippopotamus, basking in a river of melted butter amidst a breakfast-themed landscape. It features the distinctive, bulky body shape of a hippo. However, instead of the usual grey skin, the creature’s body resembles a golden-brown, crispy waffle fresh off the griddle. The skin is textured with the familiar grid pattern of a waffle, each square filled with a glistening sheen of syrup. The environment combines the natural habitat of a hippo with elements of a breakfast table setting, a river of warm, melted butter, with oversized utensils or plates peeking out from the lush, pancake-like foliage in the background, a towering pepper mill standing in for a tree.  As the sun rises in this fantastical world, it casts a warm, buttery glow over the scene. The creature, content in its butter river, lets out a yawn. Nearby, a flock of birds take flight"
+
+image = pipe(
+    prompt=prompt,
+    prompt_3=prompt_3,
+    negative_prompt="",
+    num_inference_steps=28,
+    guidance_scale=4.5,
+    max_sequence_length=512,
+).images[0]
+```
+
+## Tiny AutoEncoder for Stable Diffusion 3
+
+Tiny AutoEncoder for Stable Diffusion (TAESD3) is a tiny distilled version of Stable Diffusion 3's VAE by [Ollin Boer Bohan](https://github.com/madebyollin/taesd) that can decode [`StableDiffusion3Pipeline`] latents almost instantly.
+
+To use with Stable Diffusion 3:
+
+```python
+import torch
+from diffusers import StableDiffusion3Pipeline, AutoencoderTiny
+
+pipe = StableDiffusion3Pipeline.from_pretrained(
+    "stabilityai/stable-diffusion-3-medium-diffusers", torch_dtype=torch.float16
+)
+pipe.vae = AutoencoderTiny.from_pretrained("madebyollin/taesd3", torch_dtype=torch.float16)
+pipe = pipe.to("cuda")
+
+prompt = "slice of delicious New York-style berry cheesecake"
+image = pipe(prompt, num_inference_steps=25).images[0]
+image.save("cheesecake.png")
+```
+
 ## Loading the original checkpoints via `from_single_file`
 
 The `SD3Transformer2DModel` and `StableDiffusion3Pipeline` classes support loading the original checkpoints via the `from_single_file` method. This method allows you to load the original checkpoint files that were used to train the models.
@@ -211,17 +272,41 @@ model = SD3Transformer2DModel.from_single_file("https://huggingface.co/stability
 
 ## Loading the single checkpoint for the `StableDiffusion3Pipeline`
 
-```python
-from diffusers import StableDiffusion3Pipeline
-from transformers import T5EncoderModel
+### Loading the single file checkpoint without T5
 
-text_encoder_3 = T5EncoderModel.from_pretrained("stabilityai/stable-diffusion-3-medium-diffusers", subfolder="text_encoder_3", torch_dtype=torch.float16)
-pipe = StableDiffusion3Pipeline.from_single_file("https://huggingface.co/stabilityai/stable-diffusion-3-medium/blob/main/sd3_medium_incl_clips.safetensors", torch_dtype=torch.float16, text_encoder_3=text_encoder_3)
+```python
+import torch
+from diffusers import StableDiffusion3Pipeline
+
+pipe = StableDiffusion3Pipeline.from_single_file(
+    "https://huggingface.co/stabilityai/stable-diffusion-3-medium/blob/main/sd3_medium_incl_clips.safetensors",
+    torch_dtype=torch.float16,
+    text_encoder_3=None
+)
+pipe.enable_model_cpu_offload()
+
+image = pipe("a picture of a cat holding a sign that says hello world").images[0]
+image.save('sd3-single-file.png')
 ```
 
-<Tip>
-`from_single_file` support for the `fp8` version of the checkpoints is coming soon. Watch this space.
-</Tip>
+### Loading the single file checkpoint with T5
+
+> [!TIP]
+> The following example loads a checkpoint stored in a 8-bit floating point format which requires PyTorch 2.3 or later.
+
+```python
+import torch
+from diffusers import StableDiffusion3Pipeline
+
+pipe = StableDiffusion3Pipeline.from_single_file(
+    "https://huggingface.co/stabilityai/stable-diffusion-3-medium/blob/main/sd3_medium_incl_clips_t5xxlfp8.safetensors",
+    torch_dtype=torch.float16,
+)
+pipe.enable_model_cpu_offload()
+
+image = pipe("a picture of a cat holding a sign that says hello world").images[0]
+image.save('sd3-single-file-t5-fp8.png')
+```
 
 ## StableDiffusion3Pipeline
 

docs/source/en/api/pipelines/text_to_video_zero.md

@@ -155,28 +155,28 @@ To generate a video from prompt with additional pose control
     imageio.mimsave("video.mp4", result, fps=4)
     ```
 - #### SDXL Support
-	
+
 	Since our attention processor also works with SDXL, it can be utilized to generate a video from prompt using ControlNet models powered by SDXL:
 	```python
 	import torch
 	from diffusers import StableDiffusionXLControlNetPipeline, ControlNetModel
 	from diffusers.pipelines.text_to_video_synthesis.pipeline_text_to_video_zero import CrossFrameAttnProcessor
-	
+
 	controlnet_model_id = 'thibaud/controlnet-openpose-sdxl-1.0'
 	model_id = 'stabilityai/stable-diffusion-xl-base-1.0'
-	
+
 	controlnet = ControlNetModel.from_pretrained(controlnet_model_id, torch_dtype=torch.float16)
 	pipe = StableDiffusionControlNetPipeline.from_pretrained(
 		model_id, controlnet=controlnet, torch_dtype=torch.float16
 	).to('cuda')
-	
+
 	# Set the attention processor
 	pipe.unet.set_attn_processor(CrossFrameAttnProcessor(batch_size=2))
 	pipe.controlnet.set_attn_processor(CrossFrameAttnProcessor(batch_size=2))
-	
+
 	# fix latents for all frames
 	latents = torch.randn((1, 4, 128, 128), device="cuda", dtype=torch.float16).repeat(len(pose_images), 1, 1, 1)
-	
+
 	prompt = "Darth Vader dancing in a desert"
 	result = pipe(prompt=[prompt] * len(pose_images), image=pose_images, latents=latents).images
 	imageio.mimsave("video.mp4", result, fps=4)

docs/source/en/api/schedulers/tcd.md

@@ -10,7 +10,7 @@ an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express o
 specific language governing permissions and limitations under the License.
 -->
 
-# TCDScheduler 
+# TCDScheduler
 
 [Trajectory Consistency Distillation](https://huggingface.co/papers/2402.19159) by Jianbin Zheng, Minghui Hu, Zhongyi Fan, Chaoyue Wang, Changxing Ding, Dacheng Tao and Tat-Jen Cham introduced a Strategic Stochastic Sampling (Algorithm 4) that is capable of generating good samples in a small number of steps. Distinguishing it as an advanced iteration of the multistep scheduler (Algorithm 1) in the [Consistency Models](https://huggingface.co/papers/2303.01469), Strategic Stochastic Sampling specifically tailored for the trajectory consistency function.
 

docs/source/en/conceptual/contribution.md

@@ -22,14 +22,13 @@ We enormously value feedback from the community, so please do not be afraid to s
 
 ## Overview
 
-You can contribute in many ways ranging from answering questions on issues to adding new diffusion models to
-the core library.
+You can contribute in many ways ranging from answering questions on issues and discussions to adding new diffusion models to the core library.
 
 In the following, we give an overview of different ways to contribute, ranked by difficulty in ascending order. All of them are valuable to the community.
 
 * 1. Asking and answering questions on [the Diffusers discussion forum](https://discuss.huggingface.co/c/discussion-related-to-httpsgithubcomhuggingfacediffusers) or on [Discord](https://discord.gg/G7tWnz98XR).
-* 2. Opening new issues on [the GitHub Issues tab](https://github.com/huggingface/diffusers/issues/new/choose).
-* 3. Answering issues on [the GitHub Issues tab](https://github.com/huggingface/diffusers/issues).
+* 2. Opening new issues on [the GitHub Issues tab](https://github.com/huggingface/diffusers/issues/new/choose) or new discussions on [the GitHub Discussions tab](https://github.com/huggingface/diffusers/discussions/new/choose).
+* 3. Answering issues on [the GitHub Issues tab](https://github.com/huggingface/diffusers/issues) or discussions on [the GitHub Discussions tab](https://github.com/huggingface/diffusers/discussions).
 * 4. Fix a simple issue, marked by the "Good first issue" label, see [here](https://github.com/huggingface/diffusers/issues?q=is%3Aopen+is%3Aissue+label%3A%22good+first+issue%22).
 * 5. Contribute to the [documentation](https://github.com/huggingface/diffusers/tree/main/docs/source).
 * 6. Contribute a [Community Pipeline](https://github.com/huggingface/diffusers/issues?q=is%3Aopen+is%3Aissue+label%3Acommunity-examples).
@@ -63,7 +62,7 @@ In the same spirit, you are of immense help to the community by answering such q
 
 **Please** keep in mind that the more effort you put into asking or answering a question, the higher
 the quality of the publicly documented knowledge. In the same way, well-posed and well-answered questions create a high-quality knowledge database accessible to everybody, while badly posed questions or answers reduce the overall quality of the public knowledge database.
-In short, a high quality question or answer is *precise*, *concise*, *relevant*, *easy-to-understand*, *accessible*, and *well-formated/well-posed*. For more information, please have a look through the [How to write a good issue](#how-to-write-a-good-issue) section.
+In short, a high quality question or answer is *precise*, *concise*, *relevant*, *easy-to-understand*, *accessible*, and *well-formatted/well-posed*. For more information, please have a look through the [How to write a good issue](#how-to-write-a-good-issue) section.
 
 **NOTE about channels**:
 [*The forum*](https://discuss.huggingface.co/c/discussion-related-to-httpsgithubcomhuggingfacediffusers/63) is much better indexed by search engines, such as Google. Posts are ranked by popularity rather than chronologically. Hence, it's easier to look up questions and answers that we posted some time ago.
@@ -99,7 +98,7 @@ This means in more detail:
 - Format your code.
 - Do not include any external libraries except for Diffusers depending on them.
 - **Always** provide all necessary information about your environment; for this, you can run: `diffusers-cli env` in your shell and copy-paste the displayed information to the issue.
-- Explain the issue. If the reader doesn't know what the issue is and why it is an issue, she cannot solve it.
+- Explain the issue. If the reader doesn't know what the issue is and why it is an issue, (s)he cannot solve it.
 - **Always** make sure the reader can reproduce your issue with as little effort as possible. If your code snippet cannot be run because of missing libraries or undefined variables, the reader cannot help you. Make sure your reproducible code snippet is as minimal as possible and can be copy-pasted into a simple Python shell.
 - If in order to reproduce your issue a model and/or dataset is required, make sure the reader has access to that model or dataset. You can always upload your model or dataset to the [Hub](https://huggingface.co) to make it easily downloadable. Try to keep your model and dataset as small as possible, to make the reproduction of your issue as effortless as possible.
 
@@ -288,7 +287,7 @@ The official training examples are maintained by the Diffusers' core maintainers
 This is because of the same reasons put forward in [6. Contribute a community pipeline](#6-contribute-a-community-pipeline) for official pipelines vs. community pipelines: It is not feasible for the core maintainers to maintain all possible training methods for diffusion models.
 If the Diffusers core maintainers and the community consider a certain training paradigm to be too experimental or not popular enough, the corresponding training code should be put in the `research_projects` folder and maintained by the author.
 
-Both official training and research examples consist of a directory that contains one or more training scripts, a requirements.txt file, and a README.md file. In order for the user to make use of the
+Both official training and research examples consist of a directory that contains one or more training scripts, a `requirements.txt` file, and a `README.md` file. In order for the user to make use of the
 training examples, it is required to clone the repository:
 
 ```bash
@@ -298,7 +297,8 @@ git clone https://github.com/huggingface/diffusers
 as well as to install all additional dependencies required for training:
 
 ```bash
-pip install -r /examples/<your-example-folder>/requirements.txt
+cd diffusers
+pip install -r examples/<your-example-folder>/requirements.txt
 ```
 
 Therefore when adding an example, the `requirements.txt` file shall define all pip dependencies required for your training example so that once all those are installed, the user can run the example's training script. See, for example, the [DreamBooth `requirements.txt` file](https://github.com/huggingface/diffusers/blob/main/examples/dreambooth/requirements.txt).
@@ -316,7 +316,7 @@ Once an example script works, please make sure to add a comprehensive `README.md
 - A link to some training results (logs, models, etc.) that show what the user can expect as shown [here](https://api.wandb.ai/report/patrickvonplaten/xm6cd5q5).
 - If you are adding a non-official/research training example, **please don't forget** to add a sentence that you are maintaining this training example which includes your git handle as shown [here](https://github.com/huggingface/diffusers/tree/main/examples/research_projects/intel_opts#diffusers-examples-with-intel-optimizations).
 
-If you are contributing to the official training examples, please also make sure to add a test to [examples/test_examples.py](https://github.com/huggingface/diffusers/blob/main/examples/test_examples.py). This is not necessary for non-official training examples.
+If you are contributing to the official training examples, please also make sure to add a test to its folder such as [examples/dreambooth/test_dreambooth.py](https://github.com/huggingface/diffusers/blob/main/examples/dreambooth/test_dreambooth.py). This is not necessary for non-official training examples.
 
 ### 8. Fixing a "Good second issue"
 
@@ -418,7 +418,7 @@ You will need basic `git` proficiency to be able to contribute to
 manual. Type `git --help` in a shell and enjoy. If you prefer books, [Pro
 Git](https://git-scm.com/book/en/v2) is a very good reference.
 
-Follow these steps to start contributing ([supported Python versions](https://github.com/huggingface/diffusers/blob/main/setup.py#L244)):
+Follow these steps to start contributing ([supported Python versions](https://github.com/huggingface/diffusers/blob/83bc6c94eaeb6f7704a2a428931cf2d9ad973ae9/setup.py#L270)):
 
 1. Fork the [repository](https://github.com/huggingface/diffusers) by
 clicking on the 'Fork' button on the repository's page. This creates a copy of the code
@@ -565,4 +565,4 @@ $ git push --set-upstream origin your-branch-for-syncing
 
 ### Style guide
 
-For documentation strings, 🧨 Diffusers follows the [Google style](https://google.github.io/styleguide/pyguide.html).
+For documentation strings, 🧨 Diffusers follows the [Google style](https://google.github.io/styleguide/pyguide.html).

docs/source/en/conceptual/philosophy.md

@@ -63,7 +63,7 @@ Let's walk through more in-detail design decisions for each class.
 Pipelines are designed to be easy to use (therefore do not follow [*Simple over easy*](#simple-over-easy) 100%), are not feature complete, and should loosely be seen as examples of how to use [models](#models) and [schedulers](#schedulers) for inference.
 
 The following design principles are followed:
-- Pipelines follow the single-file policy. All pipelines can be found in individual directories under src/diffusers/pipelines. One pipeline folder corresponds to one diffusion paper/project/release. Multiple pipeline files can be gathered in one pipeline folder, as it’s done for [`src/diffusers/pipelines/stable-diffusion`](https://github.com/huggingface/diffusers/tree/main/src/diffusers/pipelines/stable_diffusion). If pipelines share similar functionality, one can make use of the [#Copied from mechanism](https://github.com/huggingface/diffusers/blob/125d783076e5bd9785beb05367a2d2566843a271/src/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion_img2img.py#L251).
+- Pipelines follow the single-file policy. All pipelines can be found in individual directories under src/diffusers/pipelines. One pipeline folder corresponds to one diffusion paper/project/release. Multiple pipeline files can be gathered in one pipeline folder, as it’s done for [`src/diffusers/pipelines/stable-diffusion`](https://github.com/huggingface/diffusers/tree/main/src/diffusers/pipelines/stable_diffusion). If pipelines share similar functionality, one can make use of the [# Copied from mechanism](https://github.com/huggingface/diffusers/blob/125d783076e5bd9785beb05367a2d2566843a271/src/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion_img2img.py#L251).
 - Pipelines all inherit from [`DiffusionPipeline`].
 - Every pipeline consists of different model and scheduler components, that are documented in the [`model_index.json` file](https://huggingface.co/runwayml/stable-diffusion-v1-5/blob/main/model_index.json), are accessible under the same name as attributes of the pipeline and can be shared between pipelines with [`DiffusionPipeline.components`](https://huggingface.co/docs/diffusers/main/en/api/diffusion_pipeline#diffusers.DiffusionPipeline.components) function.
 - Every pipeline should be loadable via the [`DiffusionPipeline.from_pretrained`](https://huggingface.co/docs/diffusers/main/en/api/diffusion_pipeline#diffusers.DiffusionPipeline.from_pretrained) function.
@@ -81,7 +81,7 @@ Models are designed as configurable toolboxes that are natural extensions of [Py
 
 The following design principles are followed:
 - Models correspond to **a type of model architecture**. *E.g.* the [`UNet2DConditionModel`] class is used for all UNet variations that expect 2D image inputs and are conditioned on some context.
-- All models can be found in [`src/diffusers/models`](https://github.com/huggingface/diffusers/tree/main/src/diffusers/models) and every model architecture shall be defined in its file, e.g. [`unet_2d_condition.py`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/unet_2d_condition.py), [`transformer_2d.py`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/transformer_2d.py), etc...
+- All models can be found in [`src/diffusers/models`](https://github.com/huggingface/diffusers/tree/main/src/diffusers/models) and every model architecture shall be defined in its file, e.g. [`unets/unet_2d_condition.py`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/unets/unet_2d_condition.py), [`transformers/transformer_2d.py`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/transformers/transformer_2d.py), etc...
 - Models **do not** follow the single-file policy and should make use of smaller model building blocks, such as [`attention.py`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/attention.py), [`resnet.py`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/resnet.py), [`embeddings.py`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/embeddings.py), etc... **Note**: This is in stark contrast to Transformers' modeling files and shows that models do not really follow the single-file policy.
 - Models intend to expose complexity, just like PyTorch's `Module` class, and give clear error messages.
 - Models all inherit from `ModelMixin` and `ConfigMixin`.
@@ -90,7 +90,7 @@ The following design principles are followed:
 - To integrate new model checkpoints whose general architecture can be classified as an architecture that already exists in Diffusers, the existing model architecture shall be adapted to make it work with the new checkpoint. One should only create a new file if the model architecture is fundamentally different.
 - Models should be designed to be easily extendable to future changes. This can be achieved by limiting public function arguments, configuration arguments, and "foreseeing" future changes, *e.g.* it is usually better to add `string` "...type" arguments that can easily be extended to new future types instead of boolean `is_..._type` arguments. Only the minimum amount of changes shall be made to existing architectures to make a new model checkpoint work.
 - The model design is a difficult trade-off between keeping code readable and concise and supporting many model checkpoints. For most parts of the modeling code, classes shall be adapted for new model checkpoints, while there are some exceptions where it is preferred to add new classes to make sure the code is kept concise and
-readable long-term, such as [UNet blocks](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/unet_2d_blocks.py) and [Attention processors](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/attention_processor.py).
+readable long-term, such as [UNet blocks](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/unets/unet_2d_blocks.py) and [Attention processors](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/attention_processor.py).
 
 ### Schedulers
 
@@ -100,11 +100,11 @@ The following design principles are followed:
 - All schedulers are found in [`src/diffusers/schedulers`](https://github.com/huggingface/diffusers/tree/main/src/diffusers/schedulers).
 - Schedulers are **not** allowed to import from large utils files and shall be kept very self-contained.
 - One scheduler Python file corresponds to one scheduler algorithm (as might be defined in a paper).
-- If schedulers share similar functionalities, we can make use of the `#Copied from` mechanism.
+- If schedulers share similar functionalities, we can make use of the `# Copied from` mechanism.
 - Schedulers all inherit from `SchedulerMixin` and `ConfigMixin`.
-- Schedulers can be easily swapped out with the [`ConfigMixin.from_config`](https://huggingface.co/docs/diffusers/main/en/api/configuration#diffusers.ConfigMixin.from_config) method as explained in detail [here](../using-diffusers/schedulers.md).
+- Schedulers can be easily swapped out with the [`ConfigMixin.from_config`](https://huggingface.co/docs/diffusers/main/en/api/configuration#diffusers.ConfigMixin.from_config) method as explained in detail [here](../using-diffusers/schedulers).
 - Every scheduler has to have a `set_num_inference_steps`, and a `step` function. `set_num_inference_steps(...)` has to be called before every denoising process, *i.e.* before `step(...)` is called.
 - Every scheduler exposes the timesteps to be "looped over" via a `timesteps` attribute, which is an array of timesteps the model will be called upon.
 - The `step(...)` function takes a predicted model output and the "current" sample (x_t) and returns the "previous", slightly more denoised sample (x_t-1).
 - Given the complexity of diffusion schedulers, the `step` function does not expose all the complexity and can be a bit of a "black box".
-- In almost all cases, novel schedulers shall be implemented in a new scheduling file.
+- In almost all cases, novel schedulers shall be implemented in a new scheduling file.

docs/source/en/training/adapt_a_model.md

@@ -26,7 +26,7 @@ pipeline.unet.config["in_channels"]
 9
 ```
 
-To adapt your text-to-image model for inpainting, you'll need to change the number of `in_channels` from 4 to 9. 
+To adapt your text-to-image model for inpainting, you'll need to change the number of `in_channels` from 4 to 9.
 
 Initialize a [`UNet2DConditionModel`] with the pretrained text-to-image model weights, and change `in_channels` to 9. Changing the number of `in_channels` means you need to set `ignore_mismatched_sizes=True` and `low_cpu_mem_usage=False` to avoid a size mismatch error because the shape is different now.
 

docs/source/en/training/controlnet.md

@@ -349,7 +349,7 @@ control_image = load_image("./conditioning_image_1.png")
 prompt = "pale golden rod circle with old lace background"
 
 generator = torch.manual_seed(0)
-image = pipe(prompt, num_inference_steps=20, generator=generator, image=control_image).images[0]
+image = pipeline(prompt, num_inference_steps=20, generator=generator, image=control_image).images[0]
 image.save("./output.png")
 ```
 
@@ -363,4 +363,4 @@ The SDXL training script is discussed in more detail in the [SDXL training](sdxl
 
 Congratulations on training your own ControlNet! To learn more about how to use your new model, the following guides may be helpful:
 
-- Learn how to [use a ControlNet](../using-diffusers/controlnet) for inference on a variety of tasks.
+- Learn how to [use a ControlNet](../using-diffusers/controlnet) for inference on a variety of tasks.

docs/source/en/training/create_dataset.md

@@ -9,7 +9,7 @@ This guide will show you two ways to create a dataset to finetune on:
 
 <Tip>
 
-💡 Learn more about how to create an image dataset for training in the [Create an image dataset](https://huggingface.co/docs/datasets/image_dataset) guide. 
+💡 Learn more about how to create an image dataset for training in the [Create an image dataset](https://huggingface.co/docs/datasets/image_dataset) guide.
 
 </Tip>
 
@@ -39,7 +39,7 @@ accelerate launch train_unconditional.py \
 
 </Tip>
 
-Start by creating a dataset with the [`ImageFolder`](https://huggingface.co/docs/datasets/image_load#imagefolder) feature, which creates an `image` column containing the PIL-encoded images. 
+Start by creating a dataset with the [`ImageFolder`](https://huggingface.co/docs/datasets/image_load#imagefolder) feature, which creates an `image` column containing the PIL-encoded images.
 
 You can use the `data_dir` or `data_files` parameters to specify the location of the dataset. The `data_files` parameter supports mapping specific files to dataset splits like `train` or `test`:
 

docs/source/en/training/distributed_inference.md

@@ -52,76 +52,6 @@ To learn more, take a look at the [Distributed Inference with 🤗 Accelerate](h
 
 </Tip>
 
-### Device placement
-
-> [!WARNING]
-> This feature is experimental and its APIs might change in the future. 
-
-With Accelerate, you can use the `device_map` to determine how to distribute the models of a pipeline across multiple devices. This is useful in situations where you have more than one GPU.
-
-For example, if you have two 8GB GPUs, then using [`~DiffusionPipeline.enable_model_cpu_offload`] may not work so well because:
-
-* it only works on a single GPU
-* a single model might not fit on a single GPU ([`~DiffusionPipeline.enable_sequential_cpu_offload`] might work but it will be extremely slow and it is also limited to a single GPU)
-
-To make use of both GPUs, you can use the "balanced" device placement strategy which splits the models across all available GPUs.
-
-> [!WARNING]
-> Only the "balanced" strategy is supported at the moment, and we plan to support additional mapping strategies in the future.
-
-```diff
-from diffusers import DiffusionPipeline
-import torch
-
-pipeline = DiffusionPipeline.from_pretrained(
--    "runwayml/stable-diffusion-v1-5", torch_dtype=torch.float16, use_safetensors=True,
-+    "runwayml/stable-diffusion-v1-5", torch_dtype=torch.float16, use_safetensors=True, device_map="balanced"
-)
-image = pipeline("a dog").images[0]
-image
-```
-
-You can also pass a dictionary to enforce the maximum GPU memory that can be used on each device:
-
-```diff
-from diffusers import DiffusionPipeline
-import torch
-
-max_memory = {0:"1GB", 1:"1GB"}
-pipeline = DiffusionPipeline.from_pretrained(
-    "runwayml/stable-diffusion-v1-5",
-    torch_dtype=torch.float16, 
-    use_safetensors=True, 
-    device_map="balanced",
-+   max_memory=max_memory
-)
-image = pipeline("a dog").images[0]
-image
-```
-
-If a device is not present in `max_memory`, then it will be completely ignored and will not participate in the device placement. 
-
-By default, Diffusers uses the maximum memory of all devices. If the models don't fit on the GPUs, they are offloaded to the CPU. If the CPU doesn't have enough memory, then you might see an error. In that case, you could defer to using [`~DiffusionPipeline.enable_sequential_cpu_offload`] and [`~DiffusionPipeline.enable_model_cpu_offload`].
-
-Call [`~DiffusionPipeline.reset_device_map`] to reset the `device_map` of a pipeline. This is also necessary if you want to use methods like `to()`, [`~DiffusionPipeline.enable_sequential_cpu_offload`], and [`~DiffusionPipeline.enable_model_cpu_offload`] on a pipeline that was device-mapped.
-
-```py
-pipeline.reset_device_map()
-```
-
-Once a pipeline has been device-mapped, you can also access its device map via `hf_device_map`:
-
-```py
-print(pipeline.hf_device_map)
-```
-
-An example device map would look like so:
-
-
-```bash
-{'unet': 1, 'vae': 1, 'safety_checker': 0, 'text_encoder': 0}
-```
-
 ## PyTorch Distributed
 
 PyTorch supports [`DistributedDataParallel`](https://pytorch.org/docs/stable/generated/torch.nn.parallel.DistributedDataParallel.html) which enables data parallelism.
@@ -176,3 +106,6 @@ Once you've completed the inference script, use the `--nproc_per_node` argument
 ```bash
 torchrun run_distributed.py --nproc_per_node=2
 ```
+
+> [!TIP]
+> You can use `device_map` within a [`DiffusionPipeline`] to distribute its model-level components on multiple devices. Refer to the [Device placement](../tutorials/inference_with_big_models#device-placement) guide to learn more.

docs/source/en/training/dreambooth.md

@@ -533,7 +533,7 @@ python train_dreambooth_lora.py \
     --resolution=256 \
     --train_batch_size=4 \
     --gradient_accumulation_steps=1 \
-    --learning_rate=1e-6 \ 
+    --learning_rate=1e-6 \
     --max_train_steps=2000 \
     --validation_prompt="a sks dog" \
     --validation_epochs=100 \

docs/source/en/training/text2image.md

@@ -181,7 +181,7 @@ accelerate launch --mixed_precision="fp16"  train_text_to_image.py \
   --max_train_steps=15000 \
   --learning_rate=1e-05 \
   --max_grad_norm=1 \
-  --enable_xformers_memory_efficient_attention
+  --enable_xformers_memory_efficient_attention \
   --lr_scheduler="constant" --lr_warmup_steps=0 \
   --output_dir="sd-naruto-model" \
   --push_to_hub

docs/source/en/tutorials/fast_diffusion.md

@@ -34,13 +34,10 @@ Install [PyTorch nightly](https://pytorch.org/) to benefit from the latest and f
 pip3 install --pre torch --index-url https://download.pytorch.org/whl/nightly/cu121
 ```
 
-<Tip>
+> [!TIP]
+> The results reported below are from a 80GB 400W A100 with its clock rate set to the maximum. 
+> If you're interested in the full benchmarking code, take a look at [huggingface/diffusion-fast](https://github.com/huggingface/diffusion-fast).
 
-The results reported below are from a 80GB 400W A100 with its clock rate set to the maximum. <br>
-
-If you're interested in the full benchmarking code, take a look at [huggingface/diffusion-fast](https://github.com/huggingface/diffusion-fast).
-
-</Tip>
 
 ## Baseline
 
@@ -170,6 +167,9 @@ Using SDPA attention and compiling both the UNet and VAE cuts the latency from 3
     <img src="https://huggingface.co/datasets/sayakpaul/sample-datasets/resolve/main/progressive-acceleration-sdxl/SDXL%2C_Batch_Size%3A_1%2C_Steps%3A_30_3.png" width=500>
 </div>
 
+> [!TIP]
+> From PyTorch 2.3.1, you can control the caching behavior of `torch.compile()`. This is particularly beneficial for compilation modes like `"max-autotune"` which performs a grid-search over several compilation flags to find the optimal configuration. Learn more in the [Compile Time Caching in torch.compile](https://pytorch.org/tutorials/recipes/torch_compile_caching_tutorial.html) tutorial. 
+
 ### Prevent graph breaks
 
 Specifying `fullgraph=True` ensures there are no graph breaks in the underlying model to take full advantage of `torch.compile` without any performance degradation. For the UNet and VAE, this means changing how you access the return variables.
@@ -222,7 +222,7 @@ First, configure all the compiler tags:
 
 ```python
 from diffusers import StableDiffusionXLPipeline
-import torch 
+import torch
 
 # Notice the two new flags at the end.
 torch._inductor.config.conv_1x1_as_mm = True

docs/source/en/tutorials/inference_with_big_models.md

@@ -0,0 +1,139 @@
+<!--Copyright 2024 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# Working with big models
+
+A modern diffusion model, like [Stable Diffusion XL (SDXL)](../using-diffusers/sdxl), is not just a single model, but a collection of multiple models. SDXL has four different model-level components:
+
+* A variational autoencoder (VAE)
+* Two text encoders
+* A UNet for denoising
+
+Usually, the text encoders and the denoiser are much larger compared to the VAE. 
+
+As models get bigger and better, it’s possible your model is so big that even a single copy won’t fit in memory. But that doesn’t mean it can’t be loaded. If you have more than one GPU, there is more memory available to store your model. In this case, it’s better to split your model checkpoint into several smaller *checkpoint shards*.
+
+When a text encoder checkpoint has multiple shards, like [T5-xxl for SD3](https://huggingface.co/stabilityai/stable-diffusion-3-medium-diffusers/tree/main/text_encoder_3), it is automatically handled by the [Transformers](https://huggingface.co/docs/transformers/index) library as it is a required dependency of Diffusers when using the [`StableDiffusion3Pipeline`]. More specifically, Transformers will automatically handle the loading of multiple shards within the requested model class and get it ready so that inference can be performed.
+
+The denoiser checkpoint can also have multiple shards and supports inference thanks to the [Accelerate](https://huggingface.co/docs/accelerate/index) library. 
+
+> [!TIP]
+> Refer to the [Handling big models for inference](https://huggingface.co/docs/accelerate/main/en/concept_guides/big_model_inference) guide for general guidance when working with big models that are hard to fit into memory.
+
+For example, let's save a sharded checkpoint for the [SDXL UNet](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/tree/main/unet):
+
+```python
+from diffusers import UNet2DConditionModel
+
+unet = UNet2DConditionModel.from_pretrained(
+    "stabilityai/stable-diffusion-xl-base-1.0", subfolder="unet"
+)
+unet.save_pretrained("sdxl-unet-sharded", max_shard_size="5GB")
+```
+
+The size of the fp32 variant of the SDXL UNet checkpoint is ~10.4GB. Set the `max_shard_size` parameter to 5GB to create 3 shards. After saving, you can load them in [`StableDiffusionXLPipeline`]:
+
+```python
+from diffusers import UNet2DConditionModel, StableDiffusionXLPipeline 
+import torch
+
+unet = UNet2DConditionModel.from_pretrained(
+    "sayakpaul/sdxl-unet-sharded", torch_dtype=torch.float16
+)
+pipeline = StableDiffusionXLPipeline.from_pretrained(
+    "stabilityai/stable-diffusion-xl-base-1.0", unet=unet, torch_dtype=torch.float16
+).to("cuda")
+
+image = pipeline("a cute dog running on the grass", num_inference_steps=30).images[0]
+image.save("dog.png")
+```
+
+If placing all the model-level components on the GPU at once is not feasible, use [`~DiffusionPipeline.enable_model_cpu_offload`] to help you: 
+
+```diff
+- pipeline.to("cuda")
++ pipeline.enable_model_cpu_offload()
+```
+
+In general, we recommend sharding when a checkpoint is more than 5GB (in fp32). 
+
+## Device placement
+
+On distributed setups, you can run inference across multiple GPUs with Accelerate.
+
+> [!WARNING]
+> This feature is experimental and its APIs might change in the future.
+
+With Accelerate, you can use the `device_map` to determine how to distribute the models of a pipeline across multiple devices. This is useful in situations where you have more than one GPU.
+
+For example, if you have two 8GB GPUs, then using [`~DiffusionPipeline.enable_model_cpu_offload`] may not work so well because:
+
+* it only works on a single GPU
+* a single model might not fit on a single GPU ([`~DiffusionPipeline.enable_sequential_cpu_offload`] might work but it will be extremely slow and it is also limited to a single GPU)
+
+To make use of both GPUs, you can use the "balanced" device placement strategy which splits the models across all available GPUs.
+
+> [!WARNING]
+> Only the "balanced" strategy is supported at the moment, and we plan to support additional mapping strategies in the future.
+
+```diff
+from diffusers import DiffusionPipeline
+import torch
+
+pipeline = DiffusionPipeline.from_pretrained(
+-    "runwayml/stable-diffusion-v1-5", torch_dtype=torch.float16, use_safetensors=True,
++    "runwayml/stable-diffusion-v1-5", torch_dtype=torch.float16, use_safetensors=True, device_map="balanced"
+)
+image = pipeline("a dog").images[0]
+image
+```
+
+You can also pass a dictionary to enforce the maximum GPU memory that can be used on each device:
+
+```diff
+from diffusers import DiffusionPipeline
+import torch
+
+max_memory = {0:"1GB", 1:"1GB"}
+pipeline = DiffusionPipeline.from_pretrained(
+    "runwayml/stable-diffusion-v1-5",
+    torch_dtype=torch.float16,
+    use_safetensors=True,
+    device_map="balanced",
++   max_memory=max_memory
+)
+image = pipeline("a dog").images[0]
+image
+```
+
+If a device is not present in `max_memory`, then it will be completely ignored and will not participate in the device placement.
+
+By default, Diffusers uses the maximum memory of all devices. If the models don't fit on the GPUs, they are offloaded to the CPU. If the CPU doesn't have enough memory, then you might see an error. In that case, you could defer to using [`~DiffusionPipeline.enable_sequential_cpu_offload`] and [`~DiffusionPipeline.enable_model_cpu_offload`].
+
+Call [`~DiffusionPipeline.reset_device_map`] to reset the `device_map` of a pipeline. This is also necessary if you want to use methods like `to()`, [`~DiffusionPipeline.enable_sequential_cpu_offload`], and [`~DiffusionPipeline.enable_model_cpu_offload`] on a pipeline that was device-mapped.
+
+```py
+pipeline.reset_device_map()
+```
+
+Once a pipeline has been device-mapped, you can also access its device map via `hf_device_map`:
+
+```py
+print(pipeline.hf_device_map)
+```
+
+An example device map would look like so:
+
+
+```bash
+{'unet': 1, 'vae': 1, 'safety_checker': 0, 'text_encoder': 0}
+```

docs/source/en/using-diffusers/controlnet.md

@@ -506,7 +506,7 @@ make_image_grid([original_image, canny_image], rows=1, cols=2)
 </div>
 
 For human pose estimation, install [controlnet_aux](https://github.com/patrickvonplaten/controlnet_aux):
-  
+
 ```py
 # uncomment to install the necessary library in Colab
 #!pip install -q controlnet-aux

docs/source/en/using-diffusers/custom_pipeline_overview.md

@@ -147,11 +147,11 @@ prompt = "cat, hiding in the leaves, ((rain)), zazie rainyday, beautiful eyes, m
 neg_prompt = "(deformed iris, deformed pupils, semi-realistic, cgi, 3d, render, sketch, cartoon, drawing, anime, mutated hands and fingers:1.4), (deformed, distorted, disfigured:1.3), poorly drawn, bad anatomy, wrong anatomy, extra limb, missing limb, floating limbs, disconnected limbs, mutation, mutated, ugly, disgusting, amputation"
 generator = torch.Generator(device="cpu").manual_seed(20)
 out_lpw = pipe_lpw(
-    prompt, 
-    negative_prompt=neg_prompt, 
+    prompt,
+    negative_prompt=neg_prompt,
     width=512,
     height=512,
-    max_embeddings_multiples=3, 
+    max_embeddings_multiples=3,
     num_inference_steps=50,
     generator=generator,
     ).images[0]

docs/source/en/using-diffusers/inference_with_lcm.md

@@ -235,7 +235,7 @@ image = pipe(
     mask_image=mask_image,
     generator=generator,
     num_inference_steps=4,
-    guidance_scale=4, 
+    guidance_scale=4,
 ).images[0]
 image
 ```
@@ -497,7 +497,7 @@ pipe = StableDiffusionXLAdapterPipeline.from_pretrained(
     unet=unet,
     adapter=adapter,
     torch_dtype=torch.float16,
-    variant="fp16", 
+    variant="fp16",
 ).to("cuda")
 
 pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)
@@ -512,7 +512,7 @@ image = pipe(
     image=canny_image,
     num_inference_steps=4,
     guidance_scale=5,
-    adapter_conditioning_scale=0.8, 
+    adapter_conditioning_scale=0.8,
     adapter_conditioning_factor=1,
     generator=generator,
 ).images[0]
@@ -554,10 +554,10 @@ canny_image = Image.fromarray(image).resize((1024, 1024))
 adapter = T2IAdapter.from_pretrained("TencentARC/t2i-adapter-canny-sdxl-1.0", torch_dtype=torch.float16, varient="fp16").to("cuda")
 
 pipe = StableDiffusionXLAdapterPipeline.from_pretrained(
-    "stabilityai/stable-diffusion-xl-base-1.0", 
+    "stabilityai/stable-diffusion-xl-base-1.0",
     adapter=adapter,
     torch_dtype=torch.float16,
-    variant="fp16", 
+    variant="fp16",
 ).to("cuda")
 
 pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)
@@ -573,8 +573,8 @@ image = pipe(
     negative_prompt=negative_prompt,
     image=canny_image,
     num_inference_steps=4,
-    guidance_scale=1.5, 
-    adapter_conditioning_scale=0.8, 
+    guidance_scale=1.5,
+    adapter_conditioning_scale=0.8,
     adapter_conditioning_factor=1,
     generator=generator,
 ).images[0]

docs/source/en/using-diffusers/ip_adapter.md

@@ -445,8 +445,8 @@ generator = torch.Generator(device="cpu").manual_seed(42)
 
 images = pipeline(
     prompt="A photo of a girl",
-    ip_adapter_image_embeds=[id_embeds], 
-    negative_prompt="monochrome, lowres, bad anatomy, worst quality, low quality", 
+    ip_adapter_image_embeds=[id_embeds],
+    negative_prompt="monochrome, lowres, bad anatomy, worst quality, low quality",
     num_inference_steps=20, num_images_per_prompt=1,
     generator=generator
 ).images
@@ -661,7 +661,7 @@ image
 
 ### Style & layout control
 
-[InstantStyle](https://arxiv.org/abs/2404.02733) is a plug-and-play method on top of IP-Adapter, which disentangles style and layout from image prompt to control image generation. This way, you can generate images following only the style or layout from image prompt, with significantly improved diversity. This is achieved by only activating IP-Adapters to specific parts of the model. 
+[InstantStyle](https://arxiv.org/abs/2404.02733) is a plug-and-play method on top of IP-Adapter, which disentangles style and layout from image prompt to control image generation. This way, you can generate images following only the style or layout from image prompt, with significantly improved diversity. This is achieved by only activating IP-Adapters to specific parts of the model.
 
 By default IP-Adapters are inserted to all layers of the model. Use the [`~loaders.IPAdapterMixin.set_ip_adapter_scale`] method with a dictionary to assign scales to IP-Adapter at different layers.
 

docs/source/en/using-diffusers/loading.md

@@ -81,14 +81,14 @@ pipeline = StableDiffusionImg2ImgPipeline.from_pretrained("runwayml/stable-diffu
 Use the Space below to gauge a pipeline's memory requirements before you download and load it to see if it runs on your hardware.
 
 <div class="block dark:hidden">
-	<iframe 
+	<iframe
         src="https://diffusers-compute-pipeline-size.hf.space?__theme=light"
         width="850"
         height="1600"
     ></iframe>
 </div>
 <div class="hidden dark:block">
-    <iframe 
+    <iframe
         src="https://diffusers-compute-pipeline-size.hf.space?__theme=dark"
         width="850"
         height="1600"
@@ -138,11 +138,11 @@ Now pass the new scheduler and VAE to the [`StableDiffusionXLPipeline`].
 
 ```py
 pipeline = StableDiffusionXLPipeline.from_pretrained(
-  "stabilityai/stable-diffusion-xl-base-1.0", 
-  scheduler=scheduler, 
-  vae=vae, 
-  torch_dtype=torch.float16, 
-  variant="fp16", 
+  "stabilityai/stable-diffusion-xl-base-1.0",
+  scheduler=scheduler,
+  vae=vae,
+  torch_dtype=torch.float16,
+  variant="fp16",
   use_safetensors=True
 ).to("cuda")
 ```
@@ -178,7 +178,7 @@ pipe_sd.to("cuda")
 generator = torch.Generator(device="cpu").manual_seed(33)
 out_sd = pipe_sd(
     prompt="bear eats pizza",
-    negative_prompt="wrong white balance, dark, sketches,worst quality,low quality", 
+    negative_prompt="wrong white balance, dark, sketches,worst quality,low quality",
     ip_adapter_image=image,
     num_inference_steps=50,
     generator=generator,
@@ -284,7 +284,7 @@ pipe.sag_unload_ip_adapter()
 generator = torch.Generator(device="cpu").manual_seed(33)
 out_sd = pipe_sd(
     prompt="bear eats pizza",
-    negative_prompt="wrong white balance, dark, sketches,worst quality,low quality", 
+    negative_prompt="wrong white balance, dark, sketches,worst quality,low quality",
     ip_adapter_image=image,
     num_inference_steps=50,
     generator=generator,

docs/source/en/using-diffusers/loading_adapters.md

@@ -166,7 +166,7 @@ For both [`~loaders.LoraLoaderMixin.load_lora_weights`] and [`~loaders.UNet2DCon
 For more granular control on the amount of LoRA weights used per layer, you can use [`~loaders.LoraLoaderMixin.set_adapters`] and pass a dictionary specifying by how much to scale the weights in each layer by.
 ```python
 pipe = ... # create pipeline
-pipe.load_lora_weights(..., adapter_name="my_adapter") 
+pipe.load_lora_weights(..., adapter_name="my_adapter")
 scales = {
     "text_encoder": 0.5,
     "text_encoder_2": 0.5,  # only usable if pipe has a 2nd text encoder

docs/source/en/using-diffusers/merge_loras.md

@@ -71,7 +71,7 @@ import torch
 
 unet = UNet2DConditionModel.from_pretrained(
     "stabilityai/stable-diffusion-xl-base-1.0",
-    torch_dtype=torch.float16, 
+    torch_dtype=torch.float16,
     use_safetensors=True,
     variant="fp16",
     subfolder="unet",
@@ -138,7 +138,7 @@ from peft import PeftModel
 
 base_unet = UNet2DConditionModel.from_pretrained(
     "stabilityai/stable-diffusion-xl-base-1.0",
-    torch_dtype=torch.float16, 
+    torch_dtype=torch.float16,
     use_safetensors=True,
     variant="fp16",
     subfolder="unet",

docs/source/en/using-diffusers/other-formats.md

@@ -418,7 +418,7 @@ my_local_checkpoint_path = hf_hub_download(
 
 my_local_config_path = snapshot_download(
     repo_id="segmind/SSD-1B",
-    allowed_patterns=["*.json", "**/*.json", "*.txt", "**/*.txt"]
+    allow_patterns=["*.json", "**/*.json", "*.txt", "**/*.txt"]
 )
 
 pipeline = StableDiffusionXLPipeline.from_single_file(my_local_checkpoint_path, config=my_local_config_path, local_files_only=True)
@@ -438,7 +438,7 @@ my_local_checkpoint_path = hf_hub_download(
 
 my_local_config_path = snapshot_download(
     repo_id="segmind/SSD-1B",
-    allowed_patterns=["*.json", "**/*.json", "*.txt", "**/*.txt"]
+    allow_patterns=["*.json", "**/*.json", "*.txt", "**/*.txt"]
     local_dir="my_local_config"
 )
 
@@ -468,11 +468,10 @@ print("My local checkpoint: ", my_local_checkpoint_path)
 
 my_local_config_path = snapshot_download(
     repo_id="segmind/SSD-1B",
-    allowed_patterns=["*.json", "**/*.json", "*.txt", "**/*.txt"]
+    allow_patterns=["*.json", "**/*.json", "*.txt", "**/*.txt"]
     local_dir_use_symlinks=False,
 )
 print("My local config: ", my_local_config_path)
-
 ```
 
 Then you can pass the local paths to the `pretrained_model_link_or_path` and `config` parameters.

docs/source/en/using-diffusers/pag.md

@@ -0,0 +1,351 @@
+<!--Copyright 2024 The HuggingFace Team. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
+an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
+specific language governing permissions and limitations under the License.
+-->
+
+# Perturbed-Attention Guidance
+
+[Perturbed-Attention Guidance (PAG)](https://ku-cvlab.github.io/Perturbed-Attention-Guidance/) is a new diffusion sampling guidance that improves sample quality across both unconditional and conditional settings, achieving this without requiring further training or the integration of external modules. PAG is designed to progressively enhance the structure of synthesized samples throughout the denoising process by considering the self-attention mechanisms' ability to capture structural information. It involves generating intermediate samples with degraded structure by substituting selected self-attention maps in diffusion U-Net with an identity matrix, and guiding the denoising process away from these degraded samples.
+
+This guide will show you how to use PAG for various tasks and use cases.
+
+
+## General tasks
+
+You can apply PAG to the [`StableDiffusionXLPipeline`] for tasks such as text-to-image, image-to-image, and inpainting. To enable PAG for a specific task, load the pipeline using the [AutoPipeline](../api/pipelines/auto_pipeline) API with the `enable_pag=True` flag and the `pag_applied_layers` argument.
+
+> [!TIP]
+> 🤗 Diffusers currently only supports using PAG with selected SDXL pipelines, but feel free to open a [feature request](https://github.com/huggingface/diffusers/issues/new/choose) if you want to add PAG support to a new pipeline!
+
+<hfoptions id="tasks">
+<hfoption id="Text-to-image">
+
+```py
+from diffusers import AutoPipelineForText2Image
+from diffusers.utils import load_image
+import torch
+
+pipeline = AutoPipelineForText2Image.from_pretrained(
+    "stabilityai/stable-diffusion-xl-base-1.0",
+    enable_pag=True,
+    pag_applied_layers=["mid"],
+    torch_dtype=torch.float16
+)
+pipeline.enable_model_cpu_offload()
+```
+
+> [!TIP]
+> The `pag_applied_layers` argument allows you to specify which layers PAG is applied to. Additionally, you can use `set_pag_applied_layers` method to update these layers after the pipeline has been created. Check out the [pag_applied_layers](#pag_applied_layers) section to learn more about applying PAG to other layers.
+
+If you already have a pipeline created and loaded, you can enable PAG on it using the `from_pipe` API with the `enable_pag` flag. Internally, a PAG pipeline is created based on the pipeline and task you specified. In the example below, since we used `AutoPipelineForText2Image` and passed a `StableDiffusionXLPipeline`, a `StableDiffusionXLPAGPipeline` is created accordingly. Note that this does not require additional memory, and you will have both `StableDiffusionXLPipeline` and  `StableDiffusionXLPAGPipeline` loaded and ready to use. You can read more about the `from_pipe` API and how to reuse pipelines in diffuser[here](https://huggingface.co/docs/diffusers/using-diffusers/loading#reuse-a-pipeline)
+
+```py
+pipeline_sdxl = AutoPipelineForText2Image.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0, torch_dtype=torch.float16")
+pipeline = AutoPipelineForText2Image.from_pipe(pipeline_sdxl, enable_pag=True)
+```
+
+To generate an image, you will also need to pass a `pag_scale`. When `pag_scale` increases, images gain more semantically coherent structures and exhibit fewer artifacts. However overly large guidance scale can lead to smoother textures and slight saturation in the images, similarly to CFG. `pag_scale=3.0` is used in the official demo and works well in most of the use cases, but feel free to experiment and select the appropriate value according to your needs! PAG is disabled when `pag_scale=0`.
+
+```py
+prompt = "an insect robot preparing a delicious meal, anime style"
+
+for pag_scale in [0.0, 3.0]:
+    generator = torch.Generator(device="cpu").manual_seed(0)
+    images = pipeline(
+        prompt=prompt,
+        num_inference_steps=25,
+        guidance_scale=7.0,
+        generator=generator,
+        pag_scale=pag_scale,
+    ).images
+```
+
+<div class="flex flex-row gap-4">
+  <div class="flex-1">
+    <img class="rounded-xl" src="https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/pag_0.0_cfg_7.0_mid.png"/>
+    <figcaption class="mt-2 text-center text-sm text-gray-500">generated image without PAG</figcaption>
+  </div>
+  <div class="flex-1">
+    <img class="rounded-xl" src="https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/pag_3.0_cfg_7.0_mid.png"/>
+    <figcaption class="mt-2 text-center text-sm text-gray-500">generated image with PAG</figcaption>
+  </div>
+</div>
+
+</hfoption>
+<hfoption id="Image-to-image">
+
+You can use PAG with image-to-image pipelines.
+
+```py
+from diffusers import AutoPipelineForImage2Image
+from diffusers.utils import load_image
+import torch
+
+pipeline = AutoPipelineForImage2Image.from_pretrained(
+    "stabilityai/stable-diffusion-xl-base-1.0",
+    enable_pag=True,
+    pag_applied_layers=["mid"],
+    torch_dtype=torch.float16
+)
+pipeline.enable_model_cpu_offload()
+```
+
+If you already have a image-to-image pipeline and would like enable PAG on it, you can run this
+
+```py
+pipeline_t2i = AutoPipelineForImage2Image.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0", torch_dtype=torch.float16)
+pipeline = AutoPipelineForImage2Image.from_pipe(pipeline_t2i, enable_pag=True)
+```
+
+It is also very easy to directly switch from a text-to-image pipeline to PAG enabled image-to-image pipeline
+
+```py
+pipeline_pag = AutoPipelineForText2Image.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0", torch_dtype=torch.float16)
+pipeline = AutoPipelineForImage2Image.from_pipe(pipeline_t2i, enable_pag=True)
+```
+
+If you have a PAG enabled text-to-image pipeline, you can directly switch to a image-to-image pipeline with PAG still enabled
+
+```py
+pipeline_pag = AutoPipelineForText2Image.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0", enable_pag=True, torch_dtype=torch.float16)
+pipeline = AutoPipelineForImage2Image.from_pipe(pipeline_t2i)
+```
+
+Now let's generate an image!
+
+```py
+pag_scales =  4.0
+guidance_scales = 7.0
+
+url = "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/sdxl-text2img.png"
+init_image = load_image(url)
+prompt = "a dog catching a frisbee in the jungle"
+
+generator = torch.Generator(device="cpu").manual_seed(0)
+image = pipeline(
+    prompt, 
+    image=init_image, 
+    strength=0.8, 
+    guidance_scale=guidance_scale, 
+    pag_scale=pag_scale,
+    generator=generator).images[0]
+```
+
+</hfoption>
+<hfoption id="Inpainting">
+
+```py
+from diffusers import AutoPipelineForInpainting
+from diffusers.utils import load_image
+import torch
+
+pipeline = AutoPipelineForInpainting.from_pretrained(
+    "stabilityai/stable-diffusion-xl-base-1.0",
+    enable_pag=True,
+    torch_dtype=torch.float16
+)
+pipeline.enable_model_cpu_offload()
+```
+
+You can enable PAG on an exisiting inpainting pipeline like this
+
+```py
+pipeline_inpaint = AutoPipelineForInpaiting.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0", torch_dtype=torch.float16)
+pipeline = AutoPipelineForInpaiting.from_pipe(pipeline_inpaint, enable_pag=True)
+```
+
+This still works when your pipeline has a different task: 
+
+```py
+pipeline_t2i = AutoPipelineForText2Image.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0", torch_dtype=torch.float16)
+pipeline = AutoPipelineForInpaiting.from_pipe(pipeline_t2i, enable_pag=True)
+```
+
+Let's generate an image! 
+
+```py
+img_url = "https://raw.githubusercontent.com/CompVis/latent-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo.png"
+mask_url = "https://raw.githubusercontent.com/CompVis/latent-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo_mask.png"
+init_image = load_image(img_url).convert("RGB")
+mask_image = load_image(mask_url).convert("RGB")
+
+prompt = "A majestic tiger sitting on a bench"
+
+pag_scales =  3.0
+guidance_scales = 7.5
+
+generator = torch.Generator(device="cpu").manual_seed(1)
+images = pipeline(
+    prompt=prompt,
+    image=init_image,
+    mask_image=mask_image,
+    strength=0.8,
+    num_inference_steps=50,
+    guidance_scale=guidance_scale,
+    generator=generator,
+    pag_scale=pag_scale,
+).images
+images[0]
+```
+</hfoption>
+</hfoptions>
+
+## PAG with ControlNet
+
+To use PAG with ControlNet, first create a `controlnet`. Then, pass the `controlnet` and other PAG arguments to the `from_pretrained` method of the AutoPipeline for the specified task.
+
+```py
+from diffusers import AutoPipelineForText2Image, ControlNetModel
+import torch
+
+controlnet = ControlNetModel.from_pretrained(
+    "diffusers/controlnet-canny-sdxl-1.0", torch_dtype=torch.float16
+)
+
+pipeline = AutoPipelineForText2Image.from_pretrained(
+    "stabilityai/stable-diffusion-xl-base-1.0",
+    controlnet=controlnet,
+    enable_pag=True,
+    pag_applied_layers="mid",
+    torch_dtype=torch.float16
+)
+pipeline.enable_model_cpu_offload()
+```
+
+<Tip>
+
+If you already have a controlnet pipeline and want to enable PAG, you can use the `from_pipe` API: `AutoPipelineForText2Image.from_pipe(pipeline_controlnet, enable_pag=True)`
+
+</Tip>
+
+You can use the pipeline in the same way you normally use ControlNet pipelines, with the added option to specify a `pag_scale` parameter. Note that PAG works well for unconditional generation. In this example, we will generate an image without a prompt.
+
+```py
+from diffusers.utils import load_image
+canny_image = load_image(
+    "https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/pag_control_input.png"
+)
+
+for pag_scale in [0.0, 3.0]:
+    generator = torch.Generator(device="cpu").manual_seed(1)
+    images = pipeline(
+        prompt="",
+        controlnet_conditioning_scale=controlnet_conditioning_scale,
+        image=canny_image,
+        num_inference_steps=50,
+        guidance_scale=0,
+        generator=generator,
+        pag_scale=pag_scale,
+    ).images
+    images[0]
+```
+
+<div class="flex flex-row gap-4">
+  <div class="flex-1">
+    <img class="rounded-xl" src="https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/pag_0.0_controlnet.png"/>
+    <figcaption class="mt-2 text-center text-sm text-gray-500">generated image without PAG</figcaption>
+  </div>
+  <div class="flex-1">
+    <img class="rounded-xl" src="https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/pag_3.0_controlnet.png"/>
+    <figcaption class="mt-2 text-center text-sm text-gray-500">generated image with PAG</figcaption>
+  </div>
+</div>
+
+## PAG with IP-Adapter 
+
+[IP-Adapter](https://hf.co/papers/2308.06721) is a popular model that can be plugged into diffusion models to enable image prompting without any changes to the underlying model. You can enable PAG on a pipeline with IP-Adapter loaded.
+
+```py
+from diffusers import AutoPipelineForText2Image
+from diffusers.utils import load_image
+from transformers import CLIPVisionModelWithProjection
+import torch
+
+image_encoder = CLIPVisionModelWithProjection.from_pretrained(
+    "h94/IP-Adapter",
+    subfolder="models/image_encoder",
+    torch_dtype=torch.float16
+)
+
+pipeline = AutoPipelineForText2Image.from_pretrained(
+    "stabilityai/stable-diffusion-xl-base-1.0",
+    image_encoder=image_encoder,
+    enable_pag=True,
+    torch_dtype=torch.float16
+).to("cuda")
+
+pipeline.load_ip_adapter("h94/IP-Adapter", subfolder="sdxl_models", weight_name="ip-adapter-plus_sdxl_vit-h.bin")
+
+pag_scales = 5.0
+ip_adapter_scales = 0.8
+
+image = load_image("https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_diner.png")
+
+pipeline.set_ip_adapter_scale(ip_adapter_scale)
+generator = torch.Generator(device="cpu").manual_seed(0)
+images = pipeline(
+    prompt="a polar bear sitting in a chair drinking a milkshake",
+    ip_adapter_image=image,
+    negative_prompt="deformed, ugly, wrong proportion, low res, bad anatomy, worst quality, low quality",
+    num_inference_steps=25,
+    guidance_scale=3.0,
+    generator=generator,
+    pag_scale=pag_scale,
+).images
+images[0]
+
+```
+
+PAG reduces artifacts and improves the overall compposition.
+
+<div class="flex flex-row gap-4">
+  <div class="flex-1">
+    <img class="rounded-xl" src="https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/pag_0.0_ipa_0.8.png"/>
+    <figcaption class="mt-2 text-center text-sm text-gray-500">generated image without PAG</figcaption>
+  </div>
+  <div class="flex-1">
+    <img class="rounded-xl" src="https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/pag_5.0_ipa_0.8.png"/>
+    <figcaption class="mt-2 text-center text-sm text-gray-500">generated image with PAG</figcaption>
+  </div>
+</div>
+
+
+## Configure parameters 
+
+### pag_applied_layers
+
+The `pag_applied_layers` argument allows you to specify which layers PAG is applied to. By default, it applies only to the mid blocks. Changing this setting will significantly impact the output. You can use the `set_pag_applied_layers` method to adjust the PAG layers after the pipeline is created, helping you find the optimal layers for your model.
+
+As an example, here is the images generated with `pag_layers = ["down.block_2"]` and `pag_layers = ["down.block_2", "up.block_1.attentions_0"]`
+
+```py
+prompt = "an insect robot preparing a delicious meal, anime style"
+pipeline.set_pag_applied_layers(pag_layers)
+generator = torch.Generator(device="cpu").manual_seed(0)
+images = pipeline(
+    prompt=prompt,
+    num_inference_steps=25,
+    guidance_scale=guidance_scale,
+    generator=generator,
+    pag_scale=pag_scale,
+).images
+images[0]
+```
+
+<div class="flex flex-row gap-4">
+  <div class="flex-1">
+    <img class="rounded-xl" src="https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/pag_3.0_cfg_7.0_down2_up1a0.png"/>
+    <figcaption class="mt-2 text-center text-sm text-gray-500">down.block_2 + up.block1.attentions_0</figcaption>
+  </div>
+  <div class="flex-1">
+    <img class="rounded-xl" src="https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/pag_3.0_cfg_7.0_down2.png"/>
+    <figcaption class="mt-2 text-center text-sm text-gray-500">down.block_2</figcaption>
+  </div>
+</div>

docs/source/en/using-diffusers/schedulers.md

@@ -186,7 +186,7 @@ scheduler, scheduler_state = FlaxDPMSolverMultistepScheduler.from_pretrained(
 pipeline, params = FlaxStableDiffusionPipeline.from_pretrained(
     "runwayml/stable-diffusion-v1-5",
     scheduler=scheduler,
-    revision="bf16",
+    variant="bf16",
     dtype=jax.numpy.bfloat16,
 )
 params["scheduler"] = scheduler_state

docs/source/en/using-diffusers/sdxl.md

@@ -63,7 +63,7 @@ from diffusers import StableDiffusionXLPipeline, StableDiffusionXLImg2ImgPipelin
 import torch
 
 pipeline = StableDiffusionXLPipeline.from_single_file(
-    "https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/blob/main/sd_xl_base_1.0.safetensors", 
+    "https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/blob/main/sd_xl_base_1.0.safetensors",
     torch_dtype=torch.float16
 ).to("cuda")
 
@@ -285,6 +285,12 @@ refiner = DiffusionPipeline.from_pretrained(
 ).to("cuda")
 ```
 
+<Tip warning={true}>
+
+You can use SDXL refiner with a different base model. For example, you can use the [Hunyuan-DiT](../../api/pipelines/hunyuandit) or [PixArt-Sigma](../../api/pipelines/pixart_sigma) pipelines to generate images with better prompt adherence. Once you have generated an image, you can pass it to the SDXL refiner model to enhance final generation quality.
+
+</Tip>
+
 Generate an image from the base model, and set the model output to **latent** space:
 
 ```py

docs/source/en/using-diffusers/stable_diffusion_jax_how_to.md

@@ -63,7 +63,7 @@ Flax is a functional framework, so models are stateless and parameters are store
 dtype = jnp.bfloat16
 pipeline, params = FlaxStableDiffusionPipeline.from_pretrained(
     "CompVis/stable-diffusion-v1-4",
-    revision="bf16",
+    variant="bf16",
     dtype=dtype,
 )
 ```

docs/source/ja/stable_diffusion.md

@@ -155,7 +155,7 @@ def get_inputs(batch_size=1):
 `batch_size=4`で開始し、どれだけメモリを消費したかを確認します：
 
 ```python
-from diffusers.utils import make_image_grid 
+from diffusers.utils import make_image_grid
 
 images = pipeline(**get_inputs(batch_size=4)).images
 make_image_grid(images, 2, 2)

docs/source/ko/conceptual/contribution.md

@@ -22,7 +22,7 @@ specific language governing permissions and limitations under the License.
 
 ## 개요
 
-이슈에 있는 질문에 답변하는 것에서부터 코어 라이브러리에 새로운 diffusion 모델을 추가하는 것까지 다양한 방법으로 기여를 할 수 있습니다. 
+이슈에 있는 질문에 답변하는 것에서부터 코어 라이브러리에 새로운 diffusion 모델을 추가하는 것까지 다양한 방법으로 기여를 할 수 있습니다.
 
 이어지는 부분에서 우리는 다양한 방법의 기여에 대한 개요를 난이도에 따라 오름차순으로 정리하였습니다. 모든 기여는 커뮤니티에게 가치가 있습니다.
 
@@ -94,7 +94,7 @@ Diffusers 라이브러리와 관련된 모든 질문이나 의견은 [토론 포
 - 이슈를 설명해야 합니다. 독자가 문제가 무엇이며 왜 문제인지 모르면 해결할 수 없습니다.
 - **항상** 독자가 가능한 한 적은 노력으로 문제를 재현할 수 있도록 해야 합니다. 코드 조각이 라이브러리가 없거나 정의되지 않은 변수 때문에 실행되지 않는 경우 독자가 도움을 줄 수 없습니다. 재현 가능한 코드 조각이 가능한 한 최소화되고 간단한 Python 셸에 복사하여 붙여넣을 수 있도록 해야 합니다.
 - 문제를 재현하기 위해 모델과/또는 데이터셋이 필요한 경우 독자가 해당 모델이나 데이터셋에 접근할 수 있도록 해야 합니다. 모델이나 데이터셋을 [Hub](https://huggingface.co)에 업로드하여 쉽게 다운로드할 수 있도록 할 수 있습니다. 문제 재현을 가능한 한 쉽게하기 위해 모델과 데이터셋을 가능한 한 작게 유지하려고 노력하세요.
-  
+
 자세한 내용은 [좋은 이슈 작성 방법](#how-to-write-a-good-issue) 섹션을 참조하세요.
 
 버그 보고서를 열려면 [여기](https://github.com/huggingface/diffusers/issues/new?assignees=&labels=bug&projects=&template=bug-report.yml)를 클릭하세요.
@@ -151,7 +151,7 @@ GitHub에서 이슈에 대한 답변을 하기 위해서는 Diffusers에 대한
 
 이슈가 올바른 버그 보고서이고 소스 코드에서 수정이 필요하다고 확인한 경우, 다음 섹션을 살펴보세요.
 
-다음 모든 기여에 대해서는 PR을 열여야 합니다. [PR 열기](#how-to-open-a-pr) 섹션에서 자세히 설명되어 있습니다. 
+다음 모든 기여에 대해서는 PR을 열여야 합니다. [PR 열기](#how-to-open-a-pr) 섹션에서 자세히 설명되어 있습니다.
 
 ### 4. "Good first issue" 고치기
 
@@ -308,7 +308,7 @@ Good second issue는 일반적으로 Good first issue 이슈보다 병합하기
 ### 9. 파이프라인, 모델, 스케줄러 추가하기
 
 파이프라인, 모델, 스케줄러는 Diffusers 라이브러리에서 가장 중요한 부분입니다.
-이들은 최첨단 diffusion 기술에 쉽게 접근하도록 하며, 따라서 커뮤니티가 강력한 생성형 AI 애플리케이션을 만들 수 있도록 합니다. 
+이들은 최첨단 diffusion 기술에 쉽게 접근하도록 하며, 따라서 커뮤니티가 강력한 생성형 AI 애플리케이션을 만들 수 있도록 합니다.
 
 새로운 모델, 파이프라인 또는 스케줄러를 추가함으로써, 사용자 인터페이스에 새로운 강력한 사용 사례를 활성화할 수 있으며, 이는 전체 생성형 AI 생태계에 매우 중요한 가치를 제공할 수 있습니다.
 

docs/source/ko/conceptual/ethical_guidelines.md

@@ -37,7 +37,7 @@ Diffusers 커뮤니티는 프로젝트의 개발에 다음과 같은 윤리 지
 
 - **간결성**: Diffusers 라이브러리를 사용하고 활용하기 쉽게 만들기 위해, 프로젝트의 목표를 간결하고 일관성 있게 유지할 것을 약속합니다.
 
-- **접근성**: Diffusers 프로젝트는 기술적 전문 지식 없어도 프로젝트 운영에 참여할 수 있는 기여자의 진입장벽을 낮춥니다. 이를 통해 연구 결과물이 커뮤니티에 더 잘 접근할 수 있게 됩니다. 
+- **접근성**: Diffusers 프로젝트는 기술적 전문 지식 없어도 프로젝트 운영에 참여할 수 있는 기여자의 진입장벽을 낮춥니다. 이를 통해 연구 결과물이 커뮤니티에 더 잘 접근할 수 있게 됩니다.
 
 - **재현성**: 우리는 Diffusers 라이브러리를 통해 제공되는 업스트림(upstream) 코드, 모델 및 데이터셋의 재현성에 대해 투명하게 공개할 것을 목표로 합니다.
 
@@ -58,6 +58,6 @@ Diffusers 커뮤니티는 프로젝트의 개발에 다음과 같은 윤리 지
 
   - [**안전 검사기**](https://github.com/huggingface/diffusers/blob/main/src/diffusers/pipelines/stable_diffusion/safety_checker.py): 이미지가 생성된 후에 이미자가 임베딩 공간에서 일련의 하드코딩된 유해 개념의 클래스일 확률을 확인하고 비교합니다. 유해 개념은 역공학을 방지하기 위해 의도적으로 숨겨져 있습니다.
 
-- **Hub에서의 단계적인 배포**: 특히 민감한 상황에서는 일부 리포지토리에 대한 접근을 제한해야 합니다. 이 단계적인 배포는 중간 단계로, 리포지토리 작성자가 사용에 대한 더 많은 통제력을 갖게 합니다. 
+- **Hub에서의 단계적인 배포**: 특히 민감한 상황에서는 일부 리포지토리에 대한 접근을 제한해야 합니다. 이 단계적인 배포는 중간 단계로, 리포지토리 작성자가 사용에 대한 더 많은 통제력을 갖게 합니다.
 
 - **라이선싱**: [OpenRAILs](https://huggingface.co/blog/open_rail)와 같은 새로운 유형의 라이선싱을 통해 자유로운 접근을 보장하면서도 더 책임 있는 사용을 위한 일련의 제한을 둘 수 있습니다.

docs/source/ko/conceptual/evaluation.md

@@ -141,7 +141,7 @@ print(images.shape)
 # (6, 512, 512, 3)
 ```
 
-그러고 나서 CLIP 점수를 계산합니다. 
+그러고 나서 CLIP 점수를 계산합니다.
 
 ```python
 from torchmetrics.functional.multimodal import clip_score

docs/source/ko/conceptual/philosophy.md

@@ -10,30 +10,30 @@ an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express o
 specific language governing permissions and limitations under the License.
 -->
 
-# 철학
+# 철학 [[philosophy]] 
 
 🧨 Diffusers는 다양한 모달리티에서 **최신의** 사전 훈련된 diffusion 모델을 제공합니다.
 그 목적은 추론과 훈련을 위한 **모듈식 툴박스**로 사용되는 것입니다.
 
-우리는 오랜 시간에 견딜 수 있는 라이브러리를 구축하는 것을 목표로 하고, 따라서 API 설계를 매우 중요시합니다.
+저희는 시간이 지나도 변치 않는 라이브러리를 구축하는 것을 목표로 하기에 API 설계를 매우 중요하게 생각합니다.
 
-간단히 말해서, Diffusers는 PyTorch의 자연스러운 확장이 되도록 구축되었습니다. 따라서 대부분의 설계 선택은 [PyTorch의 설계 원칙](https://pytorch.org/docs/stable/community/design.html#pytorch-design-philosophy)에 기반합니다. 이제 가장 중요한 것들을 살펴보겠습니다:
+간단히 말해서, Diffusers는 PyTorch를 자연스럽게 확장할 수 있도록 만들어졌습니다. 따라서 대부분의 설계 선택은 [PyTorch의 설계 원칙](https://pytorch.org/docs/stable/community/design.html#pytorch-design-philosophy)에 기반합니다. 이제 가장 중요한 것들을 살펴보겠습니다:
 
-## 성능보다는 사용성을
+## 성능보다는 사용성을 [[usability-over-performance]]
 
-- Diffusers는 많은 내장 성능 향상 기능을 갖고 있지만 (자세한 내용은 [메모리와 속도](https://huggingface.co/docs/diffusers/optimization/fp16) 참조), 모델은 항상 가장 높은 정밀도와 최소한의 최적화로 로드됩니다. 따라서 기본적인 diffusion 파이프라인은 따로 정의하지 않는다면 CPU에서 float32 정밀도로 인스턴스화됩니다. 이는 다양한 플랫폼과 가속기에서의 사용성을 보장하며, 라이브러리를 실행하기 위해 복잡한 설치가 필요하지 않음을 의미합니다.
+- Diffusers는 다양한 성능 향상 기능이 내장되어 있지만 (자세한 내용은 [메모리와 속도](https://huggingface.co/docs/diffusers/optimization/fp16) 참조), 모델은 항상 가장 높은 정밀도와 최소한의 최적화로 로드됩니다. 따라서 사용자가 별도로 정의하지 않는 한 기본적으로 diffusion 파이프라인은 항상 float32 정밀도로 CPU에 인스턴스화됩니다. 이는 다양한 플랫폼과 가속기에서의 사용성을 보장하며, 라이브러리를 실행하기 위해 복잡한 설치가 필요하지 않다는 것을 의미합니다.
 - Diffusers는 **가벼운** 패키지를 지향하기 때문에 필수 종속성은 거의 없지만 성능을 향상시킬 수 있는 많은 선택적 종속성이 있습니다 (`accelerate`, `safetensors`, `onnx` 등). 저희는 라이브러리를 가능한 한 가볍게 유지하여 다른 패키지에 대한 종속성 걱정이 없도록 노력하고 있습니다.
 - Diffusers는 간결하고 이해하기 쉬운 코드를 선호합니다. 이는 람다 함수나 고급 PyTorch 연산자와 같은 압축된 코드 구문을 자주 사용하지 않는 것을 의미합니다.
 
-## 쉬움보다는 간단함을
+## 쉬움보다는 간단함을 [[simple-over-easy]]
 
 PyTorch에서는 **명시적인 것이 암시적인 것보다 낫다**와 **단순한 것이 복잡한 것보다 낫다**라고 말합니다. 이 설계 철학은 라이브러리의 여러 부분에 반영되어 있습니다:
-- [`DiffusionPipeline.to`](https://huggingface.co/docs/diffusers/main/en/api/diffusion_pipeline#diffusers.DiffusionPipeline.to)와 같은 메서드를 사용하여 사용자가 장치 관리를 할 수 있도록 PyTorch의 API를 따릅니다.
+- [`DiffusionPipeline.to`](https://huggingface.co/docs/diffusers/main/en/api/diffusion_pipeline#diffusers.DiffusionPipeline.to)와 같은 메소드를 사용하여 사용자가 장치 관리를 할 수 있도록 PyTorch의 API를 따릅니다.
 - 잘못된 입력을 조용히 수정하는 대신 간결한 오류 메시지를 발생시키는 것이 우선입니다. Diffusers는 라이브러리를 가능한 한 쉽게 사용할 수 있도록 하는 것보다 사용자를 가르치는 것을 목표로 합니다.
 - 복잡한 모델과 스케줄러 로직이 내부에서 마법처럼 처리하는 대신 노출됩니다. 스케줄러/샘플러는 서로에게 최소한의 종속성을 가지고 분리되어 있습니다. 이로써 사용자는 언롤된 노이즈 제거 루프를 작성해야 합니다. 그러나 이 분리는 디버깅을 더 쉽게하고 노이즈 제거 과정을 조정하거나 diffusers 모델이나 스케줄러를 교체하는 데 사용자에게 더 많은 제어권을 제공합니다.
 - diffusers 파이프라인의 따로 훈련된 구성 요소인 text encoder, unet 및 variational autoencoder는 각각 자체 모델 클래스를 갖습니다. 이로써 사용자는 서로 다른 모델의 구성 요소 간의 상호 작용을 처리해야 하며, 직렬화 형식은 모델 구성 요소를 다른 파일로 분리합니다. 그러나 이는 디버깅과 커스터마이징을 더 쉽게합니다. DreamBooth나 Textual Inversion 훈련은 Diffusers의 'diffusion 파이프라인의 단일 구성 요소들을 분리할 수 있는 능력' 덕분에 매우 간단합니다.
 
-## 추상화보다는 수정 가능하고 기여하기 쉬움을
+## 추상화보다는 수정 가능하고 기여하기 쉬움을 [[tweakable-contributor-friendly-over-abstraction]]
 
 라이브러리의 대부분에 대해 Diffusers는 [Transformers 라이브러리](https://github.com/huggingface/transformers)의 중요한 설계 원칙을 채택합니다, 바로 성급한 추상화보다는 copy-pasted 코드를 선호한다는 것입니다. 이 설계 원칙은 [Don't repeat yourself (DRY)](https://en.wikipedia.org/wiki/Don%27t_repeat_yourself)와 같은 인기 있는 설계 원칙과는 대조적으로 매우 의견이 분분한데요.
 간단히 말해서, Transformers가 모델링 파일에 대해 수행하는 것처럼, Diffusers는 매우 낮은 수준의 추상화와 매우 독립적인 코드를 유지하는 것을 선호합니다. 함수, 긴 코드 블록, 심지어 클래스도 여러 파일에 복사할 수 있으며, 이는 처음에는 라이브러리를 유지할 수 없게 만드는 나쁜, 서투른 설계 선택으로 보일 수 있습니다. 하지만 이러한 설계는 매우 성공적이며, 커뮤니티 기반의 오픈 소스 기계 학습 라이브러리에 매우 적합합니다. 그 이유는 다음과 같습니다:
@@ -48,16 +48,16 @@ Diffusers에서는 이러한 철학을 파이프라인과 스케줄러에 모두
 좋아요, 이제 🧨 Diffusers가 설계된 방식을 대략적으로 이해했을 것입니다 🤗.
 우리는 이러한 설계 원칙을 일관되게 라이브러리 전체에 적용하려고 노력하고 있습니다. 그럼에도 불구하고 철학에 대한 일부 예외 사항이나 불행한 설계 선택이 있을 수 있습니다. 디자인에 대한 피드백이 있다면 [GitHub에서 직접](https://github.com/huggingface/diffusers/issues/new?assignees=&labels=&template=feedback.md&title=) 알려주시면 감사하겠습니다.
 
-## 디자인 철학 자세히 알아보기
+## 디자인 철학 자세히 알아보기 [[design-philosophy-in-details]]
 
 이제 디자인 철학의 세부 사항을 좀 더 자세히 살펴보겠습니다. Diffusers는 주로 세 가지 주요 클래스로 구성됩니다: [파이프라인](https://github.com/huggingface/diffusers/tree/main/src/diffusers/pipelines), [모델](https://github.com/huggingface/diffusers/tree/main/src/diffusers/models), 그리고 [스케줄러](https://github.com/huggingface/diffusers/tree/main/src/diffusers/schedulers). 각 클래스에 대한 더 자세한 설계 결정 사항을 살펴보겠습니다.
 
-### 파이프라인
+### 파이프라인 [[pipelines]]
 
 파이프라인은 사용하기 쉽도록 설계되었으며 (따라서 [*쉬움보다는 간단함을*](#쉬움보다는-간단함을)을 100% 따르지는 않음), feature-complete하지 않으며, 추론을 위한 [모델](#모델)과 [스케줄러](#스케줄러)를 사용하는 방법의 예시로 간주될 수 있습니다.
 
 다음과 같은 설계 원칙을 따릅니다:
-- 파이프라인은 단일 파일 정책을 따릅니다. 모든 파이프라인은 src/diffusers/pipelines의 개별 디렉토리에 있습니다. 하나의 파이프라인 폴더는 하나의 diffusion 논문/프로젝트/릴리스에 해당합니다. 여러 파이프라인 파일은 하나의 파이프라인 폴더에 모을 수 있습니다. 예를 들어 [`src/diffusers/pipelines/stable-diffusion`](https://github.com/huggingface/diffusers/tree/main/src/diffusers/pipelines/stable_diffusion)에서 그렇게 하고 있습니다. 파이프라인이 유사한 기능을 공유하는 경우, [#Copied from mechanism](https://github.com/huggingface/diffusers/blob/125d783076e5bd9785beb05367a2d2566843a271/src/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion_img2img.py#L251)을 사용할 수 있습니다.
+- 파이프라인은 단일 파일 정책을 따릅니다. 모든 파이프라인은 src/diffusers/pipelines의 개별 디렉토리에 있습니다. 하나의 파이프라인 폴더는 하나의 diffusion 논문/프로젝트/릴리스에 해당합니다. 여러 파이프라인 파일은 하나의 파이프라인 폴더에 모을 수 있습니다. 예를 들어 [`src/diffusers/pipelines/stable-diffusion`](https://github.com/huggingface/diffusers/tree/main/src/diffusers/pipelines/stable_diffusion)에서 그렇게 하고 있습니다. 파이프라인이 유사한 기능을 공유하는 경우, [# Copied from mechanism](https://github.com/huggingface/diffusers/blob/125d783076e5bd9785beb05367a2d2566843a271/src/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion_img2img.py#L251)을 사용할 수 있습니다.
 - 파이프라인은 모두 [`DiffusionPipeline`]을 상속합니다.
 - 각 파이프라인은 서로 다른 모델 및 스케줄러 구성 요소로 구성되어 있으며, 이는 [`model_index.json` 파일](https://huggingface.co/runwayml/stable-diffusion-v1-5/blob/main/model_index.json)에 문서화되어 있으며, 파이프라인의 속성 이름과 동일한 이름으로 액세스할 수 있으며, [`DiffusionPipeline.components`](https://huggingface.co/docs/diffusers/main/en/api/diffusion_pipeline#diffusers.DiffusionPipeline.components) 함수를 통해 파이프라인 간에 공유할 수 있습니다.
 - 각 파이프라인은 [`DiffusionPipeline.from_pretrained`](https://huggingface.co/docs/diffusers/main/en/api/diffusion_pipeline#diffusers.DiffusionPipeline.from_pretrained) 함수를 통해 로드할 수 있어야 합니다.
@@ -65,11 +65,11 @@ Diffusers에서는 이러한 철학을 파이프라인과 스케줄러에 모두
 - 파이프라인은 매우 가독성이 좋고, 이해하기 쉽고, 쉽게 조정할 수 있도록 설계되어야 합니다.
 - 파이프라인은 서로 상호작용하고, 상위 수준 API에 쉽게 통합할 수 있도록 설계되어야 합니다.
 - 파이프라인은 사용자 인터페이스가 feature-complete하지 않게 하는 것을 목표로 합니다. future-complete한 사용자 인터페이스를 원한다면 [InvokeAI](https://github.com/invoke-ai/InvokeAI), [Diffuzers](https://github.com/abhishekkrthakur/diffuzers), [lama-cleaner](https://github.com/Sanster/lama-cleaner)를 참조해야 합니다.
-- 모든 파이프라인은 오로지 `__call__` 메서드를 통해 실행할 수 있어야 합니다. `__call__` 인자의 이름은 모든 파이프라인에서 공유되어야 합니다.
+- 모든 파이프라인은 오로지 `__call__` 메소드를 통해 실행할 수 있어야 합니다. `__call__` 인자의 이름은 모든 파이프라인에서 공유되어야 합니다.
 - 파이프라인은 해결하고자 하는 작업의 이름으로 지정되어야 합니다.
 - 대부분의 경우에 새로운 diffusion 파이프라인은 새로운 파이프라인 폴더/파일에 구현되어야 합니다.
 
-### 모델
+### 모델 [[models]]
 
 모델은 [PyTorch의 Module 클래스](https://pytorch.org/docs/stable/generated/torch.nn.Module.html)의 자연스러운 확장이 되도록, 구성 가능한 툴박스로 설계되었습니다. 그리고 모델은 **단일 파일 정책**을 일부만 따릅니다.
 
@@ -85,7 +85,7 @@ Diffusers에서는 이러한 철학을 파이프라인과 스케줄러에 모두
 - 모델은 미래의 변경 사항을 쉽게 확장할 수 있도록 설계되어야 합니다. 이는 공개 함수 인수들과 구성 인수들을 제한하고,미래의 변경 사항을 "예상"하는 것을 통해 달성할 수 있습니다. 예를 들어, 불리언 `is_..._type` 인수보다는 새로운 미래 유형에 쉽게 확장할 수 있는 문자열 "...type" 인수를 추가하는 것이 일반적으로 더 좋습니다. 새로운 모델 체크포인트가 작동하도록 하기 위해 기존 아키텍처에 최소한의 변경만을 가해야 합니다.
 - 모델 디자인은 코드의 가독성과 간결성을 유지하는 것과 많은 모델 체크포인트를 지원하는 것 사이의 어려운 균형 조절입니다. 모델링 코드의 대부분은 새로운 모델 체크포인트를 위해 클래스를 수정하는 것이 좋지만, [UNet 블록](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/unet_2d_blocks.py) 및 [Attention 프로세서](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/attention_processor.py)와 같이 코드를 장기적으로 간결하고 읽기 쉽게 유지하기 위해 새로운 클래스를 추가하는 예외도 있습니다.
 
-### 스케줄러
+### 스케줄러 [[schedulers]]
 
 스케줄러는 추론을 위한 노이즈 제거 과정을 안내하고 훈련을 위한 노이즈 스케줄을 정의하는 역할을 합니다. 스케줄러는 개별 클래스로 설계되어 있으며, 로드 가능한 구성 파일과 **단일 파일 정책**을 엄격히 따릅니다.
 
@@ -93,9 +93,9 @@ Diffusers에서는 이러한 철학을 파이프라인과 스케줄러에 모두
 - 모든 스케줄러는 [`src/diffusers/schedulers`](https://github.com/huggingface/diffusers/tree/main/src/diffusers/schedulers)에서 찾을 수 있습니다.
 - 스케줄러는 큰 유틸리티 파일에서 가져오지 **않아야** 하며, 자체 포함성을 유지해야 합니다.
 - 하나의 스케줄러 Python 파일은 하나의 스케줄러 알고리즘(논문에서 정의된 것과 같은)에 해당합니다.
-- 스케줄러가 유사한 기능을 공유하는 경우, `#Copied from` 메커니즘을 사용할 수 있습니다.
+- 스케줄러가 유사한 기능을 공유하는 경우, `# Copied from` 메커니즘을 사용할 수 있습니다.
 - 모든 스케줄러는 `SchedulerMixin`과 `ConfigMixin`을 상속합니다.
-- [`ConfigMixin.from_config`](https://huggingface.co/docs/diffusers/main/en/api/configuration#diffusers.ConfigMixin.from_config) 메서드를 사용하여 스케줄러를 쉽게 교체할 수 있습니다. 자세한 내용은 [여기](../using-diffusers/schedulers.md)에서 설명합니다.
+- [`ConfigMixin.from_config`](https://huggingface.co/docs/diffusers/main/en/api/configuration#diffusers.ConfigMixin.from_config) 메소드를 사용하여 스케줄러를 쉽게 교체할 수 있습니다. 자세한 내용은 [여기](../using-diffusers/schedulers.md)에서 설명합니다.
 - 모든 스케줄러는 `set_num_inference_steps`와 `step` 함수를 가져야 합니다. `set_num_inference_steps(...)`는 각 노이즈 제거 과정(즉, `step(...)`이 호출되기 전) 이전에 호출되어야 합니다.
 - 각 스케줄러는 모델이 호출될 타임스텝의 배열인 `timesteps` 속성을 통해 루프를 돌 수 있는 타임스텝을 노출합니다.
 - `step(...)` 함수는 예측된 모델 출력과 "현재" 샘플(x_t)을 입력으로 받고, "이전" 약간 더 노이즈가 제거된 샘플(x_t-1)을 반환합니다.

docs/source/ko/optimization/coreml.md

@@ -158,7 +158,7 @@ Core ML 모델과 추론 코드는 🧨 Diffusers의 많은 기능, 옵션 및
 - Swift에 포팅된 스케줄러는 Stable Diffusion에서 사용하는 기본 스케줄러와 `diffusers` 구현에서 Swift로 포팅한 `DPMSolverMultistepScheduler` 두 개뿐입니다. 이들 중 약 절반의 스텝으로 동일한 품질을 생성하는 `DPMSolverMultistepScheduler`를 사용하는 것이 좋습니다.
 - 추론 코드에서 네거티브 프롬프트, classifier-free guidance scale 및 image-to-image 작업을 사용할 수 있습니다. depth guidance, ControlNet, latent upscalers와 같은 고급 기능은 아직 사용할 수 없습니다.
 
-Apple의 [변환 및 추론 리포지토리](https://github.com/apple/ml-stable-diffusion)와 자체 [swift-coreml-diffusers](https://github.com/huggingface/swift-coreml-diffusers) 리포지토리는 다른 개발자들이 구축할 수 있는 기술적인 데모입니다. 
+Apple의 [변환 및 추론 리포지토리](https://github.com/apple/ml-stable-diffusion)와 자체 [swift-coreml-diffusers](https://github.com/huggingface/swift-coreml-diffusers) 리포지토리는 다른 개발자들이 구축할 수 있는 기술적인 데모입니다.
 
 누락된 기능이 있다고 생각되면 언제든지 기능을 요청하거나, 더 좋은 방법은 기여 PR을 열어주세요. :)
 

docs/source/ko/optimization/habana.md

@@ -58,7 +58,7 @@ outputs = pipeline(
 )
 ```
 
-더 많은 정보를 얻기 위해, Optimum Habana의 [문서](https://huggingface.co/docs/optimum/habana/usage_guides/stable_diffusion)와 공식 Github 저장소에 제공된 [예시](https://github.com/huggingface/optimum-habana/tree/main/examples/stable-diffusion)를 확인하세요.
+더 많은 정보를 얻기 위해, Optimum Habana의 [문서](https://huggingface.co/docs/optimum/habana/usage_guides/stable_diffusion)와 공식 GitHub 저장소에 제공된 [예시](https://github.com/huggingface/optimum-habana/tree/main/examples/stable-diffusion)를 확인하세요.
 
 
 ## 벤치마크

docs/source/ko/optimization/opt_overview.md

@@ -12,6 +12,6 @@ specific language governing permissions and limitations under the License.
 
 # 개요
 
-노이즈가 많은 출력에서 적은 출력으로 만드는 과정으로 고품질 생성 모델의 출력을 만드는 각각의 반복되는 스텝은 많은 계산이 필요합니다. 🧨 Diffuser의 목표 중 하나는 모든 사람이 이 기술을 널리 이용할 수 있도록 하는 것이며, 여기에는 소비자 및 특수 하드웨어에서 빠른 추론을 가능하게 하는 것을 포함합니다. 
+노이즈가 많은 출력에서 적은 출력으로 만드는 과정으로 고품질 생성 모델의 출력을 만드는 각각의 반복되는 스텝은 많은 계산이 필요합니다. 🧨 Diffuser의 목표 중 하나는 모든 사람이 이 기술을 널리 이용할 수 있도록 하는 것이며, 여기에는 소비자 및 특수 하드웨어에서 빠른 추론을 가능하게 하는 것을 포함합니다.
 
 이 섹션에서는 추론 속도를 최적화하고 메모리 소비를 줄이기 위한 반정밀(half-precision) 가중치 및 sliced attention과 같은 팁과 요령을 다룹니다. 또한 [`torch.compile`](https://pytorch.org/tutorials/intermediate/torch_compile_tutorial.html) 또는 [ONNX Runtime](https://onnxruntime.ai/docs/)을 사용하여 PyTorch 코드의 속도를 높이고, [xFormers](https://facebookresearch.github.io/xformers/)를 사용하여 memory-efficient attention을 활성화하는 방법을 배울 수 있습니다. Apple Silicon, Intel 또는 Habana 프로세서와 같은 특정 하드웨어에서 추론을 실행하기 위한 가이드도 있습니다.

docs/source/ko/optimization/tome.md

@@ -96,17 +96,17 @@ We benchmarked the impact of using `tomesd` on [`StableDiffusionPipeline`] along
 |  | 2 | OOM | 13 | 10.78 |  |  |
 |  | 1 | OOM | 6.66 | 5.54 |  |  |
 
-위의 표에서 볼 수 있듯이, 이미지 해상도가 높을수록 `tomesd`를 사용한 속도 향상이 더욱 두드러집니다. 또한 `tomesd`를 사용하면 1024x1024와 같은 더 높은 해상도에서 파이프라인을 실행할 수 있다는 점도 흥미롭습니다. 
+위의 표에서 볼 수 있듯이, 이미지 해상도가 높을수록 `tomesd`를 사용한 속도 향상이 더욱 두드러집니다. 또한 `tomesd`를 사용하면 1024x1024와 같은 더 높은 해상도에서 파이프라인을 실행할 수 있다는 점도 흥미롭습니다.
 
-[`torch.compile()`](https://huggingface.co/docs/diffusers/optimization/torch2.0)을 사용하면 추론 속도를 더욱 높일 수 있습니다. 
+[`torch.compile()`](https://huggingface.co/docs/diffusers/optimization/torch2.0)을 사용하면 추론 속도를 더욱 높일 수 있습니다.
 
 ## 품질
 
-As reported in [the paper](https://arxiv.org/abs/2303.17604), ToMe can preserve the quality of the generated images to a great extent while speeding up inference. By increasing the `ratio`, it is possible to further speed up inference, but that might come at the cost of a deterioration in the image quality. 
+As reported in [the paper](https://arxiv.org/abs/2303.17604), ToMe can preserve the quality of the generated images to a great extent while speeding up inference. By increasing the `ratio`, it is possible to further speed up inference, but that might come at the cost of a deterioration in the image quality.
 
 To test the quality of the generated samples using our setup, we sampled a few prompts from the “Parti Prompts” (introduced in [Parti](https://parti.research.google/)) and performed inference with the [`StableDiffusionPipeline`] in the following settings:
 
-[논문](https://arxiv.org/abs/2303.17604)에 보고된 바와 같이, ToMe는 생성된 이미지의 품질을 상당 부분 보존하면서 추론 속도를 높일 수 있습니다. `ratio`을 높이면 추론 속도를 더 높일 수 있지만, 이미지 품질이 저하될 수 있습니다. 
+[논문](https://arxiv.org/abs/2303.17604)에 보고된 바와 같이, ToMe는 생성된 이미지의 품질을 상당 부분 보존하면서 추론 속도를 높일 수 있습니다. `ratio`을 높이면 추론 속도를 더 높일 수 있지만, 이미지 품질이 저하될 수 있습니다.
 
 해당 설정을 사용하여 생성된 샘플의 품질을 테스트하기 위해, "Parti 프롬프트"([Parti](https://parti.research.google/)에서 소개)에서 몇 가지 프롬프트를 샘플링하고 다음 설정에서 [`StableDiffusionPipeline`]을 사용하여 추론을 수행했습니다:
 
@@ -114,7 +114,7 @@ To test the quality of the generated samples using our setup, we sampled a few p
 - [`StableDiffusionPipeline`] + ToMe
 - [`StableDiffusionPipeline`] + ToMe + xformers
 
-생성된 샘플의 품질이 크게 저하되는 것을 발견하지 못했습니다. 다음은 샘플입니다: 
+생성된 샘플의 품질이 크게 저하되는 것을 발견하지 못했습니다. 다음은 샘플입니다:
 
 ![tome-samples](https://huggingface.co/datasets/diffusers/docs-images/resolve/main/tome/tome_samples.png)
 

docs/source/ko/optimization/torch2.0.md

@@ -30,7 +30,7 @@ pip install --upgrade torch diffusers
 1. **가속화된 트랜스포머 구현**
 
    PyTorch 2.0에는 [`torch.nn.functional.scaled_dot_product_attention`](https://pytorch.org/docs/master/generated/torch.nn.functional.scaled_dot_product_attention) 함수를 통해 최적화된 memory-efficient attention의 구현이 포함되어 있습니다. 이는 입력 및 GPU 유형에 따라 여러 최적화를 자동으로 활성화합니다. 이는 [xFormers](https://github.com/facebookresearch/xformers)의 `memory_efficient_attention`과 유사하지만 기본적으로 PyTorch에 내장되어 있습니다.
-   
+
    이러한 최적화는 PyTorch 2.0이 설치되어 있고 `torch.nn.functional.scaled_dot_product_attention`을 사용할 수 있는 경우 Diffusers에서 기본적으로 활성화됩니다. 이를 사용하려면 `torch 2.0`을 설치하고 파이프라인을 사용하기만 하면 됩니다. 예를 들어:
 
     ```Python
@@ -84,7 +84,7 @@ pip install --upgrade torch diffusers
     ```
 
     GPU 유형에 따라 `compile()`은 가속화된 트랜스포머 최적화를 통해 **5% - 300%**의 _추가 성능 향상_을 얻을 수 있습니다. 그러나 컴파일은 Ampere(A100, 3090), Ada(4090) 및 Hopper(H100)와 같은 최신 GPU 아키텍처에서 더 많은 성능 향상을 가져올 수 있음을 참고하세요.
-    
+
     컴파일은 완료하는 데 약간의 시간이 걸리므로, 파이프라인을 한 번 준비한 다음 동일한 유형의 추론 작업을 여러 번 수행해야 하는 상황에 가장 적합합니다. 다른 이미지 크기에서 컴파일된 파이프라인을 호출하면 시간적 비용이 많이 들 수 있는 컴파일 작업이 다시 트리거됩니다.
 
 
@@ -94,9 +94,9 @@ PyTorch 2.0의 효율적인 어텐션 구현과 `torch.compile`을 사용하여
 
 ### 벤치마킹 코드
 
-#### Stable Diffusion text-to-image 
+#### Stable Diffusion text-to-image
 
-```python 
+```python
 from diffusers import DiffusionPipeline
 import torch
 
@@ -118,9 +118,9 @@ for _ in range(3):
     images = pipe(prompt=prompt).images
 ```
 
-#### Stable Diffusion image-to-image 
+#### Stable Diffusion image-to-image
 
-```python 
+```python
 from diffusers import StableDiffusionImg2ImgPipeline
 import requests
 import torch
@@ -153,7 +153,7 @@ for _ in range(3):
 
 #### Stable Diffusion - inpainting
 
-```python 
+```python
 from diffusers import StableDiffusionInpaintPipeline
 import requests
 import torch
@@ -191,9 +191,9 @@ for _ in range(3):
     image = pipe(prompt=prompt, image=init_image, mask_image=mask_image).images[0]
 ```
 
-#### ControlNet 
+#### ControlNet
 
-```python 
+```python
 from diffusers import StableDiffusionControlNetPipeline, ControlNetModel
 import requests
 import torch
@@ -231,7 +231,7 @@ for _ in range(3):
 
 #### IF text-to-image + upscaling
 
-```python 
+```python
 from diffusers import DiffusionPipeline
 import torch
 
@@ -269,7 +269,7 @@ PyTorch 2.0 및 `torch.compile()`로 얻을 수 있는 가능한 속도 향상
 
 ![t2i_speedup](https://huggingface.co/datasets/diffusers/docs-images/resolve/main/pt2_benchmarks/t2i_speedup.png)
 
-To give you an even better idea of how this speed-up holds for the other pipelines presented above, consider the following 
+To give you an even better idea of how this speed-up holds for the other pipelines presented above, consider the following
 plot that shows the benchmarking numbers from an A100 across three different batch sizes
 (with PyTorch 2.0 nightly and `torch.compile()`):
 이 속도 향상이 위에 제시된 다른 파이프라인에 대해서도 어떻게 유지되는지 더 잘 이해하기 위해, 세 가지의 다른 배치 크기에 걸쳐 A100의 벤치마킹(PyTorch 2.0 nightly 및 `torch.compile() 사용) 수치를 보여주는 차트를 보입니다:
@@ -434,8 +434,8 @@ _(위 차트의 벤치마크 메트릭은 **초당 iteration 수(iterations/seco
 
 ## 참고
 
-* Follow [this PR](https://github.com/huggingface/diffusers/pull/3313) for more details on the environment used for conducting the benchmarks. 
-* For the IF pipeline and batch sizes > 1, we only used a batch size of >1 in the first IF pipeline for text-to-image generation and NOT for upscaling. So, that means the two upscaling pipelines received a batch size of 1. 
+* Follow [this PR](https://github.com/huggingface/diffusers/pull/3313) for more details on the environment used for conducting the benchmarks.
+* For the IF pipeline and batch sizes > 1, we only used a batch size of >1 in the first IF pipeline for text-to-image generation and NOT for upscaling. So, that means the two upscaling pipelines received a batch size of 1.
 
 *Thanks to [Horace He](https://github.com/Chillee) from the PyTorch team for their support in improving our support of `torch.compile()` in Diffusers.*
 

docs/source/ko/quicktour.md

@@ -250,7 +250,7 @@ DDPMScheduler {
 >>> less_noisy_sample.shape
 ```
 
-`less_noisy_sample`을 다음 `timestep`으로 넘기면 노이즈가 더 줄어듭니다! 이제 이 모든 것을 한데 모아 전체 노이즈 제거 과정을 시각화해 보겠습니다. 
+`less_noisy_sample`을 다음 `timestep`으로 넘기면 노이즈가 더 줄어듭니다! 이제 이 모든 것을 한데 모아 전체 노이즈 제거 과정을 시각화해 보겠습니다.
 
 먼저 노이즈 제거된 이미지를 후처리하여 `PIL.Image`로 표시하는 함수를 만듭니다:
 

docs/source/ko/stable_diffusion.md

@@ -9,7 +9,7 @@ Unless required by applicable law or agreed to in writing, software distributed
 an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
 specific language governing permissions and limitations under the License.
 -->
-                                                               
+
 # 효과적이고 효율적인 Diffusion
 
 [[open-in-colab]]
@@ -68,7 +68,7 @@ image
     <img src="https://huggingface.co/datasets/diffusers/docs-images/resolve/main/stable_diffusion_101/sd_101_1.png">
 </div>
 
-이 프로세스는 T4 GPU에서 약 30초가 소요되었습니다(할당된 GPU가 T4보다 나은 경우 더 빠를 수 있음). 기본적으로 [`DiffusionPipeline`]은 50개의 추론 단계에 대해 전체 `float32` 정밀도로 추론을 실행합니다. `float16`과 같은 더 낮은 정밀도로 전환하거나 추론 단계를 더 적게 실행하여 속도를 높일 수 있습니다. 
+이 프로세스는 T4 GPU에서 약 30초가 소요되었습니다(할당된 GPU가 T4보다 나은 경우 더 빠를 수 있음). 기본적으로 [`DiffusionPipeline`]은 50개의 추론 단계에 대해 전체 `float32` 정밀도로 추론을 실행합니다. `float16`과 같은 더 낮은 정밀도로 전환하거나 추론 단계를 더 적게 실행하여 속도를 높일 수 있습니다.
 
 `float16`으로 모델을 로드하고 이미지를 생성해 보겠습니다:
 
@@ -167,7 +167,7 @@ def image_grid(imgs, rows=2, cols=2):
         grid.paste(img, box=(i % cols * w, i // cols * h))
     return grid
 ```
- 
+
 `batch_size=4`부터 시작해 얼마나 많은 메모리를 소비했는지 확인합니다:
 
 ```python

docs/source/ko/training/controlnet.md

@@ -155,20 +155,20 @@ accelerate launch --mixed_precision="fp16" --multi_gpu train_controlnet.py \
 
 #### 배치 사이즈 8로 300 스텝 이후:
 
-| |  | 
+| |  |
 |-------------------|:-------------------------:|
-| | 푸른 배경과 빨간 원  | 
+| | 푸른 배경과 빨간 원  |
 ![conditioning image](https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/controlnet_training/conditioning_image_1.png) | ![푸른 배경과 빨간 원](https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/controlnet_training/red_circle_with_blue_background_300_steps.png) |
-| | 갈색 꽃 배경과 청록색 원 | 
+| | 갈색 꽃 배경과 청록색 원 |
 ![conditioning image](https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/controlnet_training/conditioning_image_2.png) | ![갈색 꽃 배경과 청록색 원](https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/controlnet_training/cyan_circle_with_brown_floral_background_300_steps.png) |
 
 #### 배치 사이즈 8로 6000 스텝 이후:
 
-| |  | 
+| |  |
 |-------------------|:-------------------------:|
-| | 푸른 배경과 빨간 원  | 
+| | 푸른 배경과 빨간 원  |
 ![conditioning image](https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/controlnet_training/conditioning_image_1.png) | ![푸른 배경과 빨간 원](https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/controlnet_training/red_circle_with_blue_background_6000_steps.png) |
-| | 갈색 꽃 배경과 청록색 원 | 
+| | 갈색 꽃 배경과 청록색 원 |
 ![conditioning image](https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/controlnet_training/conditioning_image_2.png) | ![갈색 꽃 배경과 청록색 원](https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/controlnet_training/cyan_circle_with_brown_floral_background_6000_steps.png) |
 
 ## 16GB GPU에서 학습하기

docs/source/ko/training/custom_diffusion.md

@@ -10,12 +10,12 @@ an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express o
 specific language governing permissions and limitations under the License.
 -->
 
-# 커스텀 Diffusion 학습 예제 
+# 커스텀 Diffusion 학습 예제
 
 [커스텀 Diffusion](https://arxiv.org/abs/2212.04488)은 피사체의 이미지 몇 장(4~5장)만 주어지면 Stable Diffusion처럼 text-to-image 모델을 커스터마이징하는 방법입니다.
 'train_custom_diffusion.py' 스크립트는 학습 과정을 구현하고 이를 Stable Diffusion에 맞게 조정하는 방법을 보여줍니다.
 
-이 교육 사례는 [Nupur Kumari](https://nupurkmr9.github.io/)가 제공하였습니다. (Custom Diffusion의 저자 중 한명). 
+이 교육 사례는 [Nupur Kumari](https://nupurkmr9.github.io/)가 제공하였습니다. (Custom Diffusion의 저자 중 한명).
 
 ## 로컬에서 PyTorch로 실행하기
 
@@ -44,7 +44,7 @@ cd examples/custom_diffusion
 
 ```bash
 pip install -r requirements.txt
-pip install clip-retrieval 
+pip install clip-retrieval
 ```
 
 그리고 [🤗Accelerate](https://github.com/huggingface/accelerate/) 환경을 초기화:
@@ -70,8 +70,8 @@ write_basic_config()
 
 이제 데이터셋을 가져옵니다. [여기](https://www.cs.cmu.edu/~custom-diffusion/assets/data.zip)에서 데이터셋을 다운로드하고 압축을 풉니다. 직접 데이터셋을 사용하려면 [학습용 데이터셋 생성하기](create_dataset) 가이드를 참고하세요.
 
-또한 'clip-retrieval'을 사용하여 200개의 실제 이미지를 수집하고, regularization으로서 이를 학습 데이터셋의 타겟 이미지와 결합합니다. 이렇게 하면 주어진 타겟 이미지에 대한 과적합을 방지할 수 있습니다. 다음 플래그를 사용하면 `prior_loss_weight=1.`로 `prior_preservation`, `real_prior` regularization을 활성화할 수 있습니다. 
-클래스_프롬프트`는 대상 이미지와 동일한 카테고리 이름이어야 합니다. 수집된 실제 이미지에는 `class_prompt`와 유사한 텍스트 캡션이 있습니다. 검색된 이미지는 `class_data_dir`에 저장됩니다. 생성된 이미지를 regularization으로 사용하기 위해 `real_prior`를 비활성화할 수 있습니다. 실제 이미지를 수집하려면 훈련 전에 이 명령을 먼저 사용하십시오. 
+또한 'clip-retrieval'을 사용하여 200개의 실제 이미지를 수집하고, regularization으로서 이를 학습 데이터셋의 타겟 이미지와 결합합니다. 이렇게 하면 주어진 타겟 이미지에 대한 과적합을 방지할 수 있습니다. 다음 플래그를 사용하면 `prior_loss_weight=1.`로 `prior_preservation`, `real_prior` regularization을 활성화할 수 있습니다.
+클래스_프롬프트`는 대상 이미지와 동일한 카테고리 이름이어야 합니다. 수집된 실제 이미지에는 `class_prompt`와 유사한 텍스트 캡션이 있습니다. 검색된 이미지는 `class_data_dir`에 저장됩니다. 생성된 이미지를 regularization으로 사용하기 위해 `real_prior`를 비활성화할 수 있습니다. 실제 이미지를 수집하려면 훈련 전에 이 명령을 먼저 사용하십시오.
 
 ```bash
 pip install clip-retrieval
@@ -110,7 +110,7 @@ accelerate launch train_custom_diffusion.py \
 가중치 및 편향(`wandb`)을 사용하여 실험을 추적하고 중간 결과를 저장하려면(강력히 권장합니다) 다음 단계를 따르세요:
 
 * `wandb` 설치: `pip install wandb`.
-* 로그인 : `wandb login`. 
+* 로그인 : `wandb login`.
 * 그런 다음 트레이닝을 시작하는 동안 `validation_prompt`를 지정하고 `report_to`를 `wandb`로 설정합니다. 다음과 같은 관련 인수를 구성할 수도 있습니다:
     * `num_validation_images`
     * `validation_steps`
@@ -136,7 +136,7 @@ accelerate launch train_custom_diffusion.py \
   --push_to_hub
 ```
 
-다음은 [Weights and Biases page](https://wandb.ai/sayakpaul/custom-diffusion/runs/26ghrcau)의 예시이며, 여러 학습 세부 정보와 함께 중간 결과들을 확인할 수 있습니다.  
+다음은 [Weights and Biases page](https://wandb.ai/sayakpaul/custom-diffusion/runs/26ghrcau)의 예시이며, 여러 학습 세부 정보와 함께 중간 결과들을 확인할 수 있습니다.
 
 `--push_to_hub`를 지정하면 학습된 파라미터가 허깅 페이스 허브의 리포지토리에 푸시됩니다. 다음은 [예제 리포지토리](https://huggingface.co/sayakpaul/custom-diffusion-cat)입니다.
 
@@ -144,7 +144,7 @@ accelerate launch train_custom_diffusion.py \
 
 [this](https://github.com/ShivamShrirao/diffusers/blob/main/examples/dreambooth/train_dreambooth.py)와 유사하게 각 컨셉에 대한 정보가 포함된 [json](https://github.com/adobe-research/custom-diffusion/blob/main/assets/concept_list.json) 파일을 제공합니다.
 
-실제 이미지를 수집하려면 json 파일의 각 컨셉에 대해 이 명령을 실행합니다. 
+실제 이미지를 수집하려면 json 파일의 각 컨셉에 대해 이 명령을 실행합니다.
 
 ```bash
 pip install clip-retrieval
@@ -287,7 +287,7 @@ image.save("multi-subject.png")
 
 ### 학습된 체크포인트에서 추론하기
 
-`--checkpointing_steps`  인수를 사용한 경우 학습 과정에서 저장된 전체 체크포인트 중 하나에서 추론을 수행할 수도 있습니다. 
+`--checkpointing_steps`  인수를 사용한 경우 학습 과정에서 저장된 전체 체크포인트 중 하나에서 추론을 수행할 수도 있습니다.
 
 ## Grads를 None으로 설정
 
@@ -297,4 +297,4 @@ image.save("multi-subject.png")
 
 ## 실험 결과
 
-실험에 대한 자세한 내용은 [당사 웹페이지](https://www.cs.cmu.edu/~custom-diffusion/)를 참조하세요. 
+실험에 대한 자세한 내용은 [당사 웹페이지](https://www.cs.cmu.edu/~custom-diffusion/)를 참조하세요.

docs/source/ko/training/dreambooth.md

@@ -59,7 +59,7 @@ DreamBooth 파인튜닝은 하이퍼파라미터에 매우 민감하고 과적
 
 <frameworkcontent>
 <pt>
-[몇 장의 강아지 이미지들](https://drive.google.com/drive/folders/1BO_dyz-p65qhBRRMRA4TbZ8qW4rB99JZ)로 DreamBooth를 시도해봅시다. 
+[몇 장의 강아지 이미지들](https://drive.google.com/drive/folders/1BO_dyz-p65qhBRRMRA4TbZ8qW4rB99JZ)로 DreamBooth를 시도해봅시다.
 이를 다운로드해 디렉터리에 저장한 다음 `INSTANCE_DIR` 환경 변수를 해당 경로로 설정합니다:
 
 
@@ -415,11 +415,11 @@ accelerate config
 ```
 
 환경 구성 중에 DeepSpeed를 사용할 것을 확인하세요.
-그러면 DeepSpeed stage 2, fp16 혼합 정밀도를 결합하고 모델 매개변수와 옵티마이저 상태를 모두 CPU로 오프로드하면 8GB VRAM 미만에서 학습할 수 있습니다. 
+그러면 DeepSpeed stage 2, fp16 혼합 정밀도를 결합하고 모델 매개변수와 옵티마이저 상태를 모두 CPU로 오프로드하면 8GB VRAM 미만에서 학습할 수 있습니다.
 단점은 더 많은 시스템 RAM(약 25GB)이 필요하다는 것입니다. 추가 구성 옵션은 [DeepSpeed 문서](https://huggingface.co/docs/accelerate/usage_guides/deepspeed)를 참조하세요.
 
 또한 기본 Adam 옵티마이저를 DeepSpeed의 최적화된 Adam 버전으로 변경해야 합니다.
-이는 상당한 속도 향상을 위한 Adam인 [`deepspeed.ops.adam.DeepSpeedCPUAdam`](https://deepspeed.readthedocs.io/en/latest/optimizers.html#adam-cpu)입니다. 
+이는 상당한 속도 향상을 위한 Adam인 [`deepspeed.ops.adam.DeepSpeedCPUAdam`](https://deepspeed.readthedocs.io/en/latest/optimizers.html#adam-cpu)입니다.
 `DeepSpeedCPUAdam`을 활성화하려면 시스템의 CUDA toolchain 버전이 PyTorch와 함께 설치된 것과 동일해야 합니다.
 
 8비트 옵티마이저는 현재 DeepSpeed와 호환되지 않는 것 같습니다.

docs/source/ko/training/instructpix2pix.md

@@ -10,7 +10,7 @@ an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express o
 specific language governing permissions and limitations under the License.
 -->
 
-# InstructPix2Pix 
+# InstructPix2Pix
 
 [InstructPix2Pix](https://arxiv.org/abs/2211.09800)는 text-conditioned diffusion 모델이 한 이미지에 편집을 따를 수 있도록 파인튜닝하는 방법입니다. 이 방법을 사용하여 파인튜닝된 모델은 다음을 입력으로 사용합니다:
 
@@ -139,7 +139,7 @@ accelerate launch --mixed_precision="fp16" train_instruct_pix2pix.py \
 `accelerate`는 원활한 다수의 GPU로 학습을 가능하게 합니다. `accelerate`로 분산 학습을 실행하는 [여기](https://huggingface.co/docs/accelerate/basic_tutorials/launch) 설명을 따라 해 주시기 바랍니다. 예시의 명령어 입니다:
 
 
-```bash 
+```bash
 accelerate launch --mixed_precision="fp16" --multi_gpu train_instruct_pix2pix.py \
  --pretrained_model_name_or_path=runwayml/stable-diffusion-v1-5 \
  --dataset_name=sayakpaul/instructpix2pix-1000-samples \

docs/source/ko/training/overview.md

@@ -12,13 +12,13 @@ specific language governing permissions and limitations under the License.
 
 # 🧨 Diffusers 학습 예시
 
-이번 챕터에서는 다양한 유즈케이스들에 대한 예제 코드들을 통해 어떻게하면 효과적으로 `diffusers` 라이브러리를 사용할 수 있을까에 대해 알아보도록 하겠습니다. 
+이번 챕터에서는 다양한 유즈케이스들에 대한 예제 코드들을 통해 어떻게하면 효과적으로 `diffusers` 라이브러리를 사용할 수 있을까에 대해 알아보도록 하겠습니다.
 
 **Note**: 혹시 오피셜한 예시코드를 찾고 있다면, [여기](https://github.com/huggingface/diffusers/tree/main/src/diffusers/pipelines)를 참고해보세요!
 
 여기서 다룰 예시들은 다음을 지향합니다.
 
-- **손쉬운 디펜던시 설치** (Self-contained) : 여기서 사용될 예시 코드들의 디펜던시 패키지들은 전부 `pip install` 명령어를 통해 설치 가능한 패키지들입니다. 또한 친절하게 `requirements.txt` 파일에 해당 패키지들이 명시되어 있어, `pip install -r requirements.txt`로 간편하게 해당 디펜던시들을 설치할 수 있습니다. 예시: [train_unconditional.py](https://github.com/huggingface/diffusers/blob/main/examples/unconditional_image_generation/train_unconditional.py), [requirements.txt](https://github.com/huggingface/diffusers/blob/main/examples/unconditional_image_generation/requirements.txt) 
+- **손쉬운 디펜던시 설치** (Self-contained) : 여기서 사용될 예시 코드들의 디펜던시 패키지들은 전부 `pip install` 명령어를 통해 설치 가능한 패키지들입니다. 또한 친절하게 `requirements.txt` 파일에 해당 패키지들이 명시되어 있어, `pip install -r requirements.txt`로 간편하게 해당 디펜던시들을 설치할 수 있습니다. 예시: [train_unconditional.py](https://github.com/huggingface/diffusers/blob/main/examples/unconditional_image_generation/train_unconditional.py), [requirements.txt](https://github.com/huggingface/diffusers/blob/main/examples/unconditional_image_generation/requirements.txt)
 - **손쉬운 수정** (Easy-to-tweak) : 저희는 가능하면 많은 유즈 케이스들을 제공하고자 합니다. 하지만 예시는 결국 그저 예시라는 점들 기억해주세요. 여기서 제공되는 예시코드들을 그저 단순히 복사-붙혀넣기하는 식으로는 여러분이 마주한 문제들을 손쉽게 해결할 순 없을 것입니다. 다시 말해 어느 정도는 여러분의 상황과 니즈에 맞춰 코드를 일정 부분 고쳐나가야 할 것입니다. 따라서 대부분의 학습 예시들은 데이터의 전처리 과정과 학습 과정에 대한 코드들을 함께 제공함으로써, 사용자가 니즈에 맞게 손쉬운 수정할 수 있도록 돕고 있습니다.
 - **입문자 친화적인** (Beginner-friendly) : 이번 챕터는 diffusion 모델과 `diffusers` 라이브러리에 대한 전반적인 이해를 돕기 위해 작성되었습니다. 따라서 diffusion 모델에 대한 최신 SOTA (state-of-the-art) 방법론들 가운데서도, 입문자에게는 많이 어려울 수 있다고 판단되면, 해당 방법론들은 여기서 다루지 않으려고 합니다.
 - **하나의 태스크만 포함할 것**(One-purpose-only): 여기서 다룰 예시들은 하나의 태스크만 포함하고 있어야 합니다. 물론 이미지 초해상화(super-resolution)와 이미지 보정(modification)과 같은 유사한 모델링 프로세스를 갖는 태스크들이 존재하겠지만, 하나의 예제에 하나의 태스크만을 담는 것이 더 이해하기 용이하다고 판단했기 때문입니다.
@@ -39,7 +39,7 @@ memory-efficient attention 연산을 수행하기 위해, 가능하면 [xFormers
 | Task | 🤗 Accelerate | 🤗 Datasets | Colab
 |---|---|:---:|:---:|
 | [**Unconditional Image Generation**](./unconditional_training) | ✅ | ✅ | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers/training_example.ipynb)
-| [**Text-to-Image fine-tuning**](./text2image) | ✅ | ✅ | 
+| [**Text-to-Image fine-tuning**](./text2image) | ✅ | ✅ |
 | [**Textual Inversion**](./text_inversion) | ✅ | - | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers/sd_textual_inversion_training.ipynb)
 | [**Dreambooth**](./dreambooth) | ✅ | - | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers/sd_dreambooth_training.ipynb)
 | [**Training with LoRA**](./lora) | ✅ | - | - |

docs/source/ko/training/text2image.md

@@ -93,7 +93,7 @@ accelerate launch train_text_to_image.py \
   --learning_rate=1e-05 \
   --max_grad_norm=1 \
   --lr_scheduler="constant" --lr_warmup_steps=0 \
-  --output_dir="sd-naruto-model" 
+  --output_dir="sd-naruto-model"
 ```
 
 자체 데이터셋으로 파인튜닝하려면 🤗 [Datasets](https://huggingface.co/docs/datasets/index)에서 요구하는 형식에 따라 데이터셋을 준비하세요. [데이터셋을 허브에 업로드](https://huggingface.co/docs/datasets/image_dataset#upload-dataset-to-the-hub)하거나 [파일들이 있는 로컬 폴더를 준비](https ://huggingface.co/docs/datasets/image_dataset#imagefolder)할 수 있습니다.
@@ -146,7 +146,7 @@ python train_text_to_image_flax.py \
   --max_train_steps=15000 \
   --learning_rate=1e-05 \
   --max_grad_norm=1 \
-  --output_dir="sd-naruto-model" 
+  --output_dir="sd-naruto-model"
 ```
 
 자체 데이터셋으로 파인튜닝하려면 🤗 [Datasets](https://huggingface.co/docs/datasets/index)에서 요구하는 형식에 따라 데이터셋을 준비하세요. [데이터셋을 허브에 업로드](https://huggingface.co/docs/datasets/image_dataset#upload-dataset-to-the-hub)하거나 [파일들이 있는 로컬 폴더를 준비](https ://huggingface.co/docs/datasets/image_dataset#imagefolder)할 수 있습니다.

docs/source/ko/training/text_inversion.md

@@ -96,13 +96,13 @@ snapshot_download(
 
 이제 [학습 스크립트](https://github.com/huggingface/diffusers/blob/main/examples/textual_inversion/textual_inversion.py)를 실행할 수 있습니다. 스크립트는 다음 파일을 생성하고 리포지토리에 저장합니다.
 
-- `learned_embeds.bin` 
+- `learned_embeds.bin`
 - `token_identifier.txt`
 - `type_of_concept.txt`.
 
 <Tip>
 
-💡V100 GPU 1개를 기준으로 전체 학습에는 최대 1시간이 걸립니다. 학습이 완료되기를 기다리는 동안 궁금한 점이 있으면 아래 섹션에서 [textual-inversion이 어떻게 작동하는지](https://huggingface.co/docs/diffusers/training/text_inversion#how-it-works) 자유롭게 확인하세요 ! 
+💡V100 GPU 1개를 기준으로 전체 학습에는 최대 1시간이 걸립니다. 학습이 완료되기를 기다리는 동안 궁금한 점이 있으면 아래 섹션에서 [textual-inversion이 어떻게 작동하는지](https://huggingface.co/docs/diffusers/training/text_inversion#how-it-works) 자유롭게 확인하세요 !
 
 </Tip>
 

docs/source/ko/tutorials/tutorial_overview.md

@@ -16,7 +16,7 @@ specific language governing permissions and limitations under the License.
 
 여러분은 이 튜토리얼을 통해 빠르게 생성하기 위해선 추론 파이프라인을 어떻게 사용해야 하는지, 그리고 라이브러리를 modular toolbox처럼 이용해서 여러분만의 diffusion system을 구축할 수 있도록 파이프라인을 분해하는 법을 배울 수 있습니다. 다음 단원에서는 여러분이 원하는 것을 생성하기 위해 자신만의 diffusion model을 학습하는 방법을 배우게 됩니다.
 
-튜토리얼을 완료한다면 여러분은 라이브러리를 직접 탐색하고, 자신의 프로젝트와 애플리케이션에 적용할 스킬들을 습득할 수 있을 겁니다. 
+튜토리얼을 완료한다면 여러분은 라이브러리를 직접 탐색하고, 자신의 프로젝트와 애플리케이션에 적용할 스킬들을 습득할 수 있을 겁니다.
 
 [Discord](https://discord.com/invite/JfAtkvEtRb)나 [포럼](https://discuss.huggingface.co/c/discussion-related-to-httpsgithubcomhuggingfacediffusers/63) 커뮤니티에 자유롭게 참여해서 다른 사용자와 개발자들과 교류하고 협업해 보세요!
 

docs/source/ko/using-diffusers/conditional_image_generation.md

@@ -28,7 +28,7 @@ specific language governing permissions and limitations under the License.
 >>> generator = DiffusionPipeline.from_pretrained("CompVis/ldm-text2im-large-256")
 ```
 
-[`DiffusionPipeline`]은 모든 모델링, 토큰화, 스케줄링 구성 요소를 다운로드하고 캐시합니다. 
+[`DiffusionPipeline`]은 모든 모델링, 토큰화, 스케줄링 구성 요소를 다운로드하고 캐시합니다.
 이 모델은 약 14억 개의 파라미터로 구성되어 있기 때문에 GPU에서 실행할 것을 강력히 권장합니다.
 PyTorch에서와 마찬가지로 생성기 객체를 GPU로 이동할 수 있습니다:
 

docs/source/ko/using-diffusers/contribute_pipeline.md

@@ -14,7 +14,7 @@ specific language governing permissions and limitations under the License.
 
 <Tip>
 
-💡 모든 사람이 속도 저하 없이 쉽게 작업을 공유할 수 있도록 커뮤니티 파이프라인을 추가하는 이유에 대한 자세한 내용은 GitHub 이슈 [#841](https://github.com/huggingface/diffusers/issues/841)를 참조하세요. 
+💡 모든 사람이 속도 저하 없이 쉽게 작업을 공유할 수 있도록 커뮤니티 파이프라인을 추가하는 이유에 대한 자세한 내용은 GitHub 이슈 [#841](https://github.com/huggingface/diffusers/issues/841)를 참조하세요.
 
 </Tip>
 
@@ -50,7 +50,7 @@ class UnetSchedulerOneForwardPipeline(DiffusionPipeline):
 +         self.register_modules(unet=unet, scheduler=scheduler)
 ```
 
-이제 '초기화' 단계가 완료되었으니 forward pass로 이동할 수 있습니다! 🔥 
+이제 '초기화' 단계가 완료되었으니 forward pass로 이동할 수 있습니다! 🔥
 
 ## Forward pass 정의
 

docs/source/ko/using-diffusers/controlling_generation.md

@@ -192,9 +192,9 @@ MultiDiffusion은 사전 학습된 diffusion model을 통해 새로운 생성
 
 ## Custom Diffusion
 
-[Custom Diffusion](../training/custom_diffusion)은 사전 학습된 text-to-image 간 확산 모델의 교차 관심도 맵만 미세 조정합니다. 
-또한 textual inversion을 추가로 수행할 수 있습니다. 설계상 다중 개념 훈련을 지원합니다. 
-DreamBooth 및 Textual Inversion 마찬가지로, 사용자 지정 확산은 사전학습된 text-to-image diffusion 모델에 새로운 개념을 학습시켜 관심 있는 개념과 관련된 출력을 생성하는 데에도 사용됩니다. 
+[Custom Diffusion](../training/custom_diffusion)은 사전 학습된 text-to-image 간 확산 모델의 교차 관심도 맵만 미세 조정합니다.
+또한 textual inversion을 추가로 수행할 수 있습니다. 설계상 다중 개념 훈련을 지원합니다.
+DreamBooth 및 Textual Inversion 마찬가지로, 사용자 지정 확산은 사전학습된 text-to-image diffusion 모델에 새로운 개념을 학습시켜 관심 있는 개념과 관련된 출력을 생성하는 데에도 사용됩니다.
 
 자세한 설명은 [공식 문서](../training/custom_diffusion)를 참조하세요.
 
@@ -202,7 +202,7 @@ DreamBooth 및 Textual Inversion 마찬가지로, 사용자 지정 확산은 사
 
 [Paper](https://arxiv.org/abs/2303.08084)
 
-[텍스트-이미지 모델 편집 파이프라인](../api/pipelines/model_editing)을 사용하면 사전학습된 text-to-image diffusion 모델이 입력 프롬프트에 있는 피사체에 대해 내릴 수 있는 잘못된 암시적 가정을 완화하는 데 도움이 됩니다. 
+[텍스트-이미지 모델 편집 파이프라인](../api/pipelines/model_editing)을 사용하면 사전학습된 text-to-image diffusion 모델이 입력 프롬프트에 있는 피사체에 대해 내릴 수 있는 잘못된 암시적 가정을 완화하는 데 도움이 됩니다.
 예를 들어, 안정적 확산에 "A pack of roses"에 대한 이미지를 생성하라는 메시지를 표시하면 생성된 이미지의 장미는 빨간색일 가능성이 높습니다. 이 파이프라인은 이러한 가정을 변경하는 데 도움이 됩니다.
 
 자세한 설명은 [공식 문서](../api/pipelines/model_editing)를 참조하세요.
@@ -221,6 +221,6 @@ DreamBooth 및 Textual Inversion 마찬가지로, 사용자 지정 확산은 사
 [Paper](https://arxiv.org/abs/2302.08453)
 
 [T2I-어댑터](../api/pipelines/stable_diffusion/adapter)는 추가적인 조건을 추가하는 auxiliary 네트워크입니다.
-가장자리 감지, 스케치, depth maps, semantic segmentations와 같은 다양한 조건에 대해 훈련된 8개의 표준 사전훈련된 adapter가 있습니다, 
+가장자리 감지, 스케치, depth maps, semantic segmentations와 같은 다양한 조건에 대해 훈련된 8개의 표준 사전훈련된 adapter가 있습니다,
 
 [공식 문서](api/pipelines/stable_diffusion/adapter)에서 사용 방법에 대한 정보를 참조하세요.

docs/source/ko/using-diffusers/custom_pipeline_examples.md

@@ -260,7 +260,7 @@ diffuser_pipeline = DiffusionPipeline.from_pretrained(
     custom_pipeline="speech_to_image_diffusion",
     speech_model=model,
     speech_processor=processor,
-    
+
     torch_dtype=torch.float16,
 )
 

docs/source/ko/using-diffusers/loading.md

@@ -45,7 +45,7 @@ repo_id = "runwayml/stable-diffusion-v1-5"
 pipe = DiffusionPipeline.from_pretrained(repo_id)
 ```
 
-물론 [`DiffusionPipeline`] 클래스를 사용하지 않고, 명시적으로 직접 해당 파이프라인 클래스를 불러오는 것도 가능합니다. 아래 예시 코드는 위 예시와 동일한 인스턴스를 반환합니다. 
+물론 [`DiffusionPipeline`] 클래스를 사용하지 않고, 명시적으로 직접 해당 파이프라인 클래스를 불러오는 것도 가능합니다. 아래 예시 코드는 위 예시와 동일한 인스턴스를 반환합니다.
 
 ```python
 from diffusers import StableDiffusionPipeline
@@ -89,7 +89,7 @@ stable_diffusion = DiffusionPipeline.from_pretrained(repo_id)
 
 ### 파이프라인 내부의 컴포넌트 교체하기
 
-파이프라인 내부의 컴포넌트들은 호환 가능한 다른 컴포넌트로 교체될 수 있습니다. 이와 같은 컴포넌트 교체가 중요한 이유는 다음과 같습니다. 
+파이프라인 내부의 컴포넌트들은 호환 가능한 다른 컴포넌트로 교체될 수 있습니다. 이와 같은 컴포넌트 교체가 중요한 이유는 다음과 같습니다.
 
 - 어떤 스케줄러를 사용할 것인가는 생성속도와 생성품질 간의 트레이드오프를 정의하는 중요한 요소입니다.
 - diffusion 모델 내부의 컴포넌트들은 일반적으로 각각 독립적으로 훈련되기 때문에, 더 좋은 성능을 보여주는 컴포넌트가 있다면 그걸로 교체하는 식으로 성능을 향상시킬 수 있습니다.
@@ -105,7 +105,7 @@ stable_diffusion = DiffusionPipeline.from_pretrained(repo_id)
 stable_diffusion.scheduler.compatibles
 ```
 
-이번에는 [`SchedulerMixin.from_pretrained`] 메서드를 사용해서, 기존 기본 스케줄러였던 [`PNDMScheduler`]를 보다 우수한 성능의 [`EulerDiscreteScheduler`]로 바꿔봅시다. 스케줄러를 로드할 때는 `subfolder` 인자를 통해, 해당 파이프라인의 리포지토리에서 [스케줄러에 관한 하위폴더](https://huggingface.co/runwayml/stable-diffusion-v1-5/tree/main/scheduler)를  명시해주어야 합니다. 
+이번에는 [`SchedulerMixin.from_pretrained`] 메서드를 사용해서, 기존 기본 스케줄러였던 [`PNDMScheduler`]를 보다 우수한 성능의 [`EulerDiscreteScheduler`]로 바꿔봅시다. 스케줄러를 로드할 때는 `subfolder` 인자를 통해, 해당 파이프라인의 리포지토리에서 [스케줄러에 관한 하위폴더](https://huggingface.co/runwayml/stable-diffusion-v1-5/tree/main/scheduler)를  명시해주어야 합니다.
 
 그 다음 새롭게 생성한 [`EulerDiscreteScheduler`] 인스턴스를 [`DiffusionPipeline`]의 `scheduler` 인자에 전달합니다.
 
@@ -149,7 +149,7 @@ components = stable_diffusion_txt2img.components
 stable_diffusion_img2img = StableDiffusionImg2ImgPipeline(**components)
 ```
 
-물론 각각의 컴포넌트들을 따로 따로 파이프라인에 전달할 수도 있습니다.  예를 들어 `stable_diffusion_txt2img` 파이프라인 안의 컴포넌트들 가운데서 세이프티 체커(`safety_checker`)와 피쳐 익스트랙터(`feature_extractor`)를 제외한 컴포넌트들만 `stable_diffusion_img2img` 파이프라인에서 재사용하는 방식 역시 가능합니다. 
+물론 각각의 컴포넌트들을 따로 따로 파이프라인에 전달할 수도 있습니다.  예를 들어 `stable_diffusion_txt2img` 파이프라인 안의 컴포넌트들 가운데서 세이프티 체커(`safety_checker`)와 피쳐 익스트랙터(`feature_extractor`)를 제외한 컴포넌트들만 `stable_diffusion_img2img` 파이프라인에서 재사용하는 방식 역시 가능합니다.
 
 ```python
 from diffusers import StableDiffusionPipeline, StableDiffusionImg2ImgPipeline
@@ -172,12 +172,12 @@ stable_diffusion_img2img = StableDiffusionImg2ImgPipeline(
 
 Variant란 일반적으로 다음과 같은 체크포인트들을 의미합니다.
 
--  `torch.float16`과 같이 정밀도는 더 낮지만, 용량 역시 더 작은 부동소수점 타입의 가중치를 사용하는 체크포인트. *(다만 이와 같은 variant의 경우, 추가적인 훈련과 CPU환경에서의 구동이 불가능합니다.)* 
+-  `torch.float16`과 같이 정밀도는 더 낮지만, 용량 역시 더 작은 부동소수점 타입의 가중치를 사용하는 체크포인트. *(다만 이와 같은 variant의 경우, 추가적인 훈련과 CPU환경에서의 구동이 불가능합니다.)*
 - Non-EMA 가중치를 사용하는 체크포인트. *(Non-EMA 가중치의 경우, 파인 튜닝 단계에서 사용하는 것이 권장되는데, 추론 단계에선 사용하지 않는 것이 권장됩니다.)*
 
 <Tip>
 
-💡 모델 구조는 동일하지만 서로 다른 학습 환경에서 서로 다른 데이터셋으로 학습된 체크포인트들이 있을 경우, 해당 체크포인트들은 variant 단계가 아닌 리포지토리 단계에서 분리되어 관리되어야 합니다. (즉, 해당 체크포인트들은 서로 다른 리포지토리에서 따로 관리되어야 합니다. 예시: [`stable-diffusion-v1-4`], [`stable-diffusion-v1-5`]). 
+💡 모델 구조는 동일하지만 서로 다른 학습 환경에서 서로 다른 데이터셋으로 학습된 체크포인트들이 있을 경우, 해당 체크포인트들은 variant 단계가 아닌 리포지토리 단계에서 분리되어 관리되어야 합니다. (즉, 해당 체크포인트들은 서로 다른 리포지토리에서 따로 관리되어야 합니다. 예시: [`stable-diffusion-v1-4`], [`stable-diffusion-v1-5`]).
 
 </Tip>
 
@@ -238,7 +238,7 @@ repo_id = "runwayml/stable-diffusion-v1-5"
 model = UNet2DConditionModel.from_pretrained(repo_id, subfolder="unet")
 ```
 
-혹은 [해당 모델의 리포지토리](https://huggingface.co/google/ddpm-cifar10-32/tree/main)로부터 다이렉트로 가져오는 것 역시 가능합니다. 
+혹은 [해당 모델의 리포지토리](https://huggingface.co/google/ddpm-cifar10-32/tree/main)로부터 다이렉트로 가져오는 것 역시 가능합니다.
 
 ```python
 from diffusers import UNet2DModel
@@ -247,7 +247,7 @@ repo_id = "google/ddpm-cifar10-32"
 model = UNet2DModel.from_pretrained(repo_id)
 ```
 
-또한 앞서 봤던 `variant` 인자를 명시함으로써, Non-EMA나 `fp16`의 가중치를 가져오는 것 역시 가능합니다. 
+또한 앞서 봤던 `variant` 인자를 명시함으로써, Non-EMA나 `fp16`의 가중치를 가져오는 것 역시 가능합니다.
 
 ```python
 from diffusers import UNet2DConditionModel
@@ -258,7 +258,7 @@ model.save_pretrained("./local-unet", variant="non-ema")
 
 ### 스케줄러
 
-스케줄러들은 [`SchedulerMixin.from_pretrained`] 메서드를 통해 불러올 수 있습니다. 모델과 달리 스케줄러는 별도의 가중치를 갖지 않으며, 따라서 당연히 별도의 학습과정을 요구하지 않습니다. 이러한 스케줄러들은 (해당 스케줄러 하위폴더의) configration 파일을 통해 정의됩니다. 
+스케줄러들은 [`SchedulerMixin.from_pretrained`] 메서드를 통해 불러올 수 있습니다. 모델과 달리 스케줄러는 별도의 가중치를 갖지 않으며, 따라서 당연히 별도의 학습과정을 요구하지 않습니다. 이러한 스케줄러들은 (해당 스케줄러 하위폴더의) configration 파일을 통해 정의됩니다.
 
 여러개의 스케줄러를 불러온다고 해서 많은 메모리를 소모하는 것은 아니며, 다양한 스케줄러들에 동일한 스케줄러 configration을  적용하는 것 역시 가능합니다. 다음 예시 코드에서 불러오는 스케줄러들은 모두 [`StableDiffusionPipeline`]과 호환되는데, 이는 곧 해당 스케줄러들에 동일한 스케줄러 configration 파일을 적용할 수 있음을 의미합니다.
 
@@ -376,7 +376,7 @@ StableDiffusionPipeline {
     ├── diffusion_pytorch_model.bin
 ```
 
-또한 각각의 컴포넌트들을 파이프라인 인스턴스의 속성으로써 참조할 수 있습니다. 
+또한 각각의 컴포넌트들을 파이프라인 인스턴스의 속성으로써 참조할 수 있습니다.
 
 ```py
 pipeline.tokenizer

docs/source/ko/using-diffusers/schedulers.md

@@ -30,7 +30,7 @@ diffusion 파이프라인은 diffusion 모델, 스케줄러 등의 컴포넌트
 
 ## 파이프라인 불러오기
 
-먼저 스테이블 diffusion 파이프라인을 불러오도록 해보겠습니다. 물론 스테이블 diffusion을 사용하기 위해서는, 허깅페이스 허브에 등록된 사용자여야 하며, 관련 [라이센스](https://huggingface.co/runwayml/stable-diffusion-v1-5)에 동의해야 한다는 점을 잊지 말아주세요. 
+먼저 스테이블 diffusion 파이프라인을 불러오도록 해보겠습니다. 물론 스테이블 diffusion을 사용하기 위해서는, 허깅페이스 허브에 등록된 사용자여야 하며, 관련 [라이센스](https://huggingface.co/runwayml/stable-diffusion-v1-5)에 동의해야 한다는 점을 잊지 말아주세요.
 
 *역자 주: 다만, 현재 신규로 생성한 허깅페이스 계정에 대해서는 라이센스 동의를 요구하지 않는 것으로 보입니다!*
 
@@ -58,7 +58,7 @@ pipeline.to("cuda")
 
 ## 스케줄러 액세스
 
-스케줄러는 언제나 파이프라인의 컴포넌트로서 존재하며, 일반적으로 파이프라인 인스턴스 내에 `scheduler`라는 이름의 속성(property)으로 정의되어 있습니다. 
+스케줄러는 언제나 파이프라인의 컴포넌트로서 존재하며, 일반적으로 파이프라인 인스턴스 내에 `scheduler`라는 이름의 속성(property)으로 정의되어 있습니다.
 
 ```python
 pipeline.scheduler
@@ -82,13 +82,13 @@ PNDMScheduler {
 }
 ```
 
-출력 결과를 통해, 우리는 해당 스케줄러가 [`PNDMScheduler`]의 인스턴스라는 것을 알 수 있습니다. 이제 [`PNDMScheduler`]와 다른 스케줄러들의 성능을 비교해보도록 하겠습니다. 먼저 테스트에 사용할 프롬프트를 다음과 같이 정의해보도록 하겠습니다. 
+출력 결과를 통해, 우리는 해당 스케줄러가 [`PNDMScheduler`]의 인스턴스라는 것을 알 수 있습니다. 이제 [`PNDMScheduler`]와 다른 스케줄러들의 성능을 비교해보도록 하겠습니다. 먼저 테스트에 사용할 프롬프트를 다음과 같이 정의해보도록 하겠습니다.
 
 ```python
 prompt = "A photograph of an astronaut riding a horse on Mars, high resolution, high definition."
 ```
 
-다음으로 유사한 이미지 생성을 보장하기 위해서, 다음과 같이 랜덤시드를 고정해주도록 하겠습니다. 
+다음으로 유사한 이미지 생성을 보장하기 위해서, 다음과 같이 랜덤시드를 고정해주도록 하겠습니다.
 
 ```python
 generator = torch.Generator(device="cuda").manual_seed(8)
@@ -107,7 +107,7 @@ image
 
 ## 스케줄러 교체하기
 
-다음으로 파이프라인의 스케줄러를 다른 스케줄러로 교체하는 방법에 대해 알아보겠습니다. 모든 스케줄러는 [`SchedulerMixin.compatibles`]라는 속성(property)을 갖고 있습니다. 해당 속성은 **호환 가능한** 스케줄러들에 대한 정보를 담고 있습니다. 
+다음으로 파이프라인의 스케줄러를 다른 스케줄러로 교체하는 방법에 대해 알아보겠습니다. 모든 스케줄러는 [`SchedulerMixin.compatibles`]라는 속성(property)을 갖고 있습니다. 해당 속성은 **호환 가능한** 스케줄러들에 대한 정보를 담고 있습니다.
 
 ```python
 pipeline.scheduler.compatibles
@@ -127,12 +127,12 @@ pipeline.scheduler.compatibles
 
 호환되는 스케줄러들을 살펴보면 아래와 같습니다.
 
-- [`LMSDiscreteScheduler`], 
-- [`DDIMScheduler`], 
-- [`DPMSolverMultistepScheduler`], 
-- [`EulerDiscreteScheduler`], 
-- [`PNDMScheduler`], 
-- [`DDPMScheduler`], 
+- [`LMSDiscreteScheduler`],
+- [`DDIMScheduler`],
+- [`DPMSolverMultistepScheduler`],
+- [`EulerDiscreteScheduler`],
+- [`PNDMScheduler`],
+- [`DDPMScheduler`],
 - [`EulerAncestralDiscreteScheduler`].
 
 앞서 정의했던 프롬프트를 사용해서 각각의 스케줄러들을 비교해보도록 하겠습니다.
@@ -161,7 +161,7 @@ FrozenDict([('num_train_timesteps', 1000),
             ('clip_sample', False)])
 ```
 
-기존 스케줄러의 config를 호환 가능한 다른 스케줄러에 이식하는 것 역시 가능합니다. 
+기존 스케줄러의 config를 호환 가능한 다른 스케줄러에 이식하는 것 역시 가능합니다.
 
 다음 예시는 기존 스케줄러(`pipeline.scheduler`)를 다른 종류의 스케줄러(`DDIMScheduler`)로 바꾸는 코드입니다. 기존 스케줄러가 갖고 있던 config를 `.from_config` 메서드의 인자로 전달하는 것을 확인할 수 있습니다.
 
@@ -296,7 +296,7 @@ scheduler, scheduler_state = FlaxDPMSolverMultistepScheduler.from_pretrained(
 pipeline, params = FlaxStableDiffusionPipeline.from_pretrained(
     model_id,
     scheduler=scheduler,
-    revision="bf16",
+    variant="bf16",
     dtype=jax.numpy.bfloat16,
 )
 params["scheduler"] = scheduler_state

docs/source/ko/using-diffusers/stable_diffusion_jax_how_to.md

@@ -83,7 +83,7 @@ Flax는 함수형 프레임워크이므로 모델은 무상태(stateless)형이
 ```python
 pipeline, params = FlaxStableDiffusionPipeline.from_pretrained(
     "CompVis/stable-diffusion-v1-4",
-    revision="bf16",
+    variant="bf16",
     dtype=dtype,
 )
 ```

docs/source/ko/using-diffusers/unconditional_image_generation.md

@@ -54,7 +54,7 @@ Unconditional 이미지 생성은 비교적 간단한 작업입니다. 모델이
 ```python
  >>> image.save("generated_image.png")
 ```
-	
+
 아래 스페이스(데모 링크)를 이용해 보고, 추론 단계의 매개변수를 자유롭게 조절하여 이미지 품질에 어떤 영향을 미치는지 확인해 보세요!
 
 <iframe src="https://stevhliu-ddpm-butterflies-128.hf.space" frameborder="0" width="850" height="500"></iframe>

docs/source/zh/stable_diffusion.md

@@ -9,7 +9,7 @@ Unless required by applicable law or agreed to in writing, software distributed
 an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
 specific language governing permissions and limitations under the License.
 -->
-                                                               
+
 # 有效且高效的扩散
 
 [[open-in-colab]]
@@ -97,7 +97,7 @@ image
 
 </Tip>
 
-另一个选择是减少推理步数。 你可以选择一个更高效的调度器 (*scheduler*) 可以减少推理步数同时保证输出质量。您可以在 [DiffusionPipeline] 中通过调用compatibles方法找到与当前模型兼容的调度器 (*scheduler*)。 
+另一个选择是减少推理步数。 你可以选择一个更高效的调度器 (*scheduler*) 可以减少推理步数同时保证输出质量。您可以在 [DiffusionPipeline] 中通过调用compatibles方法找到与当前模型兼容的调度器 (*scheduler*)。
 
 ```python
 pipeline.scheduler.compatibles
@@ -159,7 +159,7 @@ def get_inputs(batch_size=1):
 设置 `batch_size=4` ，然后看一看我们消耗了多少内存:
 
 ```python
-from diffusers.utils import make_image_grid 
+from diffusers.utils import make_image_grid
 
 images = pipeline(**get_inputs(batch_size=4)).images
 make_image_grid(images, 2, 2)

examples/advanced_diffusion_training/README.md

@@ -11,16 +11,16 @@ In a nutshell, LoRA allows to adapt pretrained models by adding pairs of rank-de
 - Previous pretrained weights are kept frozen so that the model is not prone to [catastrophic forgetting](https://www.pnas.org/doi/10.1073/pnas.1611835114)
 - Rank-decomposition matrices have significantly fewer parameters than the original model, which means that trained LoRA weights are easily portable.
 - LoRA attention layers allow to control to which extent the model is adapted towards new training images via a `scale` parameter.
-[cloneofsimo](https://github.com/cloneofsimo) was the first to try out LoRA training for Stable Diffusion in 
+[cloneofsimo](https://github.com/cloneofsimo) was the first to try out LoRA training for Stable Diffusion in
 the popular [lora](https://github.com/cloneofsimo/lora) GitHub repository.
 
-The `train_dreambooth_lora_sdxl_advanced.py` script shows how to implement dreambooth-LoRA, combining the training process shown in `train_dreambooth_lora_sdxl.py`, with 
-advanced features and techniques, inspired and built upon contributions by [Nataniel Ruiz](https://twitter.com/natanielruizg): [Dreambooth](https://dreambooth.github.io), [Rinon Gal](https://twitter.com/RinonGal): [Textual Inversion](https://textual-inversion.github.io), [Ron Mokady](https://twitter.com/MokadyRon): [Pivotal Tuning](https://arxiv.org/abs/2106.05744), [Simo Ryu](https://twitter.com/cloneofsimo): [cog-sdxl](https://github.com/replicate/cog-sdxl), 
+The `train_dreambooth_lora_sdxl_advanced.py` script shows how to implement dreambooth-LoRA, combining the training process shown in `train_dreambooth_lora_sdxl.py`, with
+advanced features and techniques, inspired and built upon contributions by [Nataniel Ruiz](https://twitter.com/natanielruizg): [Dreambooth](https://dreambooth.github.io), [Rinon Gal](https://twitter.com/RinonGal): [Textual Inversion](https://textual-inversion.github.io), [Ron Mokady](https://twitter.com/MokadyRon): [Pivotal Tuning](https://arxiv.org/abs/2106.05744), [Simo Ryu](https://twitter.com/cloneofsimo): [cog-sdxl](https://github.com/replicate/cog-sdxl),
 [Kohya](https://twitter.com/kohya_tech/): [sd-scripts](https://github.com/kohya-ss/sd-scripts), [The Last Ben](https://twitter.com/__TheBen): [fast-stable-diffusion](https://github.com/TheLastBen/fast-stable-diffusion) ❤️
 
 > [!NOTE]
-> 💡If this is your first time training a Dreambooth LoRA, congrats!🥳 
-> You might want to familiarize yourself more with the techniques: [Dreambooth blog](https://huggingface.co/blog/dreambooth), [Using LoRA for Efficient Stable Diffusion Fine-Tuning blog](https://huggingface.co/blog/lora) 
+> 💡If this is your first time training a Dreambooth LoRA, congrats!🥳
+> You might want to familiarize yourself more with the techniques: [Dreambooth blog](https://huggingface.co/blog/dreambooth), [Using LoRA for Efficient Stable Diffusion Fine-Tuning blog](https://huggingface.co/blog/lora)
 
 📚 Read more about the advanced features and best practices in this community derived blog post: [LoRA training scripts of the world, unite!](https://huggingface.co/blog/sdxl_lora_advanced_script)
 
@@ -64,19 +64,19 @@ from accelerate.utils import write_basic_config
 write_basic_config()
 ```
 
-When running `accelerate config`, if we specify torch compile mode to True there can be dramatic speedups. 
+When running `accelerate config`, if we specify torch compile mode to True there can be dramatic speedups.
 Note also that we use PEFT library as backend for LoRA training, make sure to have `peft>=0.6.0` installed in your environment.
 
 ### Pivotal Tuning
 **Training with text encoder(s)**
 
-Alongside the UNet, LoRA fine-tuning of the text encoders is also supported. In addition to the text encoder optimization 
+Alongside the UNet, LoRA fine-tuning of the text encoders is also supported. In addition to the text encoder optimization
 available with `train_dreambooth_lora_sdxl_advanced.py`, in the advanced script **pivotal tuning** is also supported.
-[pivotal tuning](https://huggingface.co/blog/sdxl_lora_advanced_script#pivotal-tuning) combines Textual Inversion with regular diffusion fine-tuning - 
-we insert new tokens into the text encoders of the model, instead of reusing existing ones. 
-We then optimize the newly-inserted token embeddings to represent the new concept. 
+[pivotal tuning](https://huggingface.co/blog/sdxl_lora_advanced_script#pivotal-tuning) combines Textual Inversion with regular diffusion fine-tuning -
+we insert new tokens into the text encoders of the model, instead of reusing existing ones.
+We then optimize the newly-inserted token embeddings to represent the new concept.
 
-To do so, just specify `--train_text_encoder_ti` while launching training (for regular text encoder optimizations, use `--train_text_encoder`). 
+To do so, just specify `--train_text_encoder_ti` while launching training (for regular text encoder optimizations, use `--train_text_encoder`).
 Please keep the following points in mind:
 
 * SDXL has two text encoders. So, we fine-tune both using LoRA.
@@ -101,7 +101,7 @@ snapshot_download(
 
 Let's review some of the advanced features we're going to be using for this example:
 - **custom captions**:
-To use custom captioning, first ensure that you have the datasets library installed, otherwise you can install it by 
+To use custom captioning, first ensure that you have the datasets library installed, otherwise you can install it by
 ```bash
 pip install datasets
 ```
@@ -113,11 +113,11 @@ Now we'll simply specify the name of the dataset and caption column (in this cas
 --caption_column=prompt
 ```
 
-You can also load a dataset straight from by specifying it's name in `dataset_name`. 
+You can also load a dataset straight from by specifying it's name in `dataset_name`.
 Look [here](https://huggingface.co/blog/sdxl_lora_advanced_script#custom-captioning) for more info on creating/loadin your own caption dataset.
 
 - **optimizer**: for this example, we'll use [prodigy](https://huggingface.co/blog/sdxl_lora_advanced_script#adaptive-optimizers) - an adaptive optimizer
-- **pivotal tuning** 
+- **pivotal tuning**
 - **min SNR gamma**
 
 **Now, we can launch training:**
@@ -161,7 +161,7 @@ accelerate launch train_dreambooth_lora_sdxl_advanced.py \
 To better track our training experiments, we're using the following flags in the command above:
 
 * `report_to="wandb` will ensure the training runs are tracked on Weights and Biases. To use it, be sure to install `wandb` with `pip install wandb`.
-* `validation_prompt` and `validation_epochs` to allow the script to do a few validation inference runs. This allows us to qualitatively check if the training is progressing as expected. 
+* `validation_prompt` and `validation_epochs` to allow the script to do a few validation inference runs. This allows us to qualitatively check if the training is progressing as expected.
 
 Our experiments were conducted on a single 40GB A100 GPU.
 
@@ -204,11 +204,11 @@ pipe.load_textual_inversion(state_dict["clip_l"], token=["<s0>", "<s1>"], text_e
 pipe.load_textual_inversion(state_dict["clip_g"], token=["<s0>", "<s1>"], text_encoder=pipe.text_encoder_2, tokenizer=pipe.tokenizer_2)
 ```
 
-3. let's generate images 
+3. let's generate images
 
 ```python
 instance_token = "<s0><s1>"
-prompt = f"a {instance_token} icon of an orange llama eating ramen, in the style of {instance_token}" 
+prompt = f"a {instance_token} icon of an orange llama eating ramen, in the style of {instance_token}"
 
 image = pipe(prompt=prompt, num_inference_steps=25, cross_attention_kwargs={"scale": 1.0}).images[0]
 image.save("llama.png")
@@ -218,37 +218,37 @@ image.save("llama.png")
 The new script fully supports textual inversion loading with Comfy UI and AUTOMATIC1111 formats!
 
 **AUTOMATIC1111 / SD.Next** \
-In AUTOMATIC1111/SD.Next we will load a LoRA and a textual embedding at the same time. 
-- *LoRA*: Besides the diffusers format, the script will also train a WebUI compatible LoRA. It is generated as `{your_lora_name}.safetensors`. You can then include it in your `models/Lora` directory. 
-- *Embedding*: the embedding is the same for diffusers and WebUI. You can download your `{lora_name}_emb.safetensors` file from a trained model, and include it in your `embeddings` directory. 
+In AUTOMATIC1111/SD.Next we will load a LoRA and a textual embedding at the same time.
+- *LoRA*: Besides the diffusers format, the script will also train a WebUI compatible LoRA. It is generated as `{your_lora_name}.safetensors`. You can then include it in your `models/Lora` directory.
+- *Embedding*: the embedding is the same for diffusers and WebUI. You can download your `{lora_name}_emb.safetensors` file from a trained model, and include it in your `embeddings` directory.
 
-You can then run inference by prompting `a y2k_emb webpage about the movie Mean Girls <lora:y2k:0.9>`. You can use the `y2k_emb` token normally, including increasing its weight by doing `(y2k_emb:1.2)`. 
+You can then run inference by prompting `a y2k_emb webpage about the movie Mean Girls <lora:y2k:0.9>`. You can use the `y2k_emb` token normally, including increasing its weight by doing `(y2k_emb:1.2)`.
 
 **ComfyUI** \
-In ComfyUI we will load a LoRA and a textual embedding at the same time. 
+In ComfyUI we will load a LoRA and a textual embedding at the same time.
 - *LoRA*: Besides the diffusers format, the script will also train a ComfyUI compatible LoRA. It is generated as `{your_lora_name}.safetensors`. You can then include it in your `models/Lora` directory. Then you will load the LoRALoader node and hook that up with your model and CLIP. [Official guide for loading LoRAs](https://comfyanonymous.github.io/ComfyUI_examples/lora/)
-- *Embedding*: the embedding is the same for diffusers and WebUI. You can download your `{lora_name}_emb.safetensors` file from a trained model, and include it in your `models/embeddings` directory and use it in your prompts like `embedding:y2k_emb`. [Official guide for loading embeddings](https://comfyanonymous.github.io/ComfyUI_examples/textual_inversion_embeddings/). 
-- 
+- *Embedding*: the embedding is the same for diffusers and WebUI. You can download your `{lora_name}_emb.safetensors` file from a trained model, and include it in your `models/embeddings` directory and use it in your prompts like `embedding:y2k_emb`. [Official guide for loading embeddings](https://comfyanonymous.github.io/ComfyUI_examples/textual_inversion_embeddings/).
+-
 ### Specifying a better VAE
 
 SDXL's VAE is known to suffer from numerical instability issues. This is why we also expose a CLI argument namely `--pretrained_vae_model_name_or_path` that lets you specify the location of a better VAE (such as [this one](https://huggingface.co/madebyollin/sdxl-vae-fp16-fix)).
 
-### DoRA training 
+### DoRA training
 The advanced script supports DoRA training too!
-> Proposed in [DoRA: Weight-Decomposed Low-Rank Adaptation](https://arxiv.org/abs/2402.09353), 
-**DoRA** is very similar to LoRA, except it decomposes the pre-trained weight into two components, **magnitude** and **direction** and employs LoRA for _directional_ updates to efficiently minimize the number of trainable parameters. 
-The authors found that by using DoRA, both the learning capacity and training stability of LoRA are enhanced without any additional overhead during inference. 
+> Proposed in [DoRA: Weight-Decomposed Low-Rank Adaptation](https://arxiv.org/abs/2402.09353),
+**DoRA** is very similar to LoRA, except it decomposes the pre-trained weight into two components, **magnitude** and **direction** and employs LoRA for _directional_ updates to efficiently minimize the number of trainable parameters.
+The authors found that by using DoRA, both the learning capacity and training stability of LoRA are enhanced without any additional overhead during inference.
 
 > [!NOTE]
-> 💡DoRA training is still _experimental_  
+> 💡DoRA training is still _experimental_
 > and is likely to require different hyperparameter values to perform best compared to a LoRA.
-> Specifically, we've noticed 2 differences to take into account your training: 
+> Specifically, we've noticed 2 differences to take into account your training:
 > 1. **LoRA seem to converge faster than DoRA** (so a set of parameters that may lead to overfitting when training a LoRA may be working well for a DoRA)
-> 2. **DoRA quality superior to LoRA especially in lower ranks** the difference in quality of DoRA of rank 8 and LoRA of rank 8 appears to be more significant than when training ranks of 32 or 64 for example.  
-> This is also aligned with some of the quantitative analysis shown in the paper. 
+> 2. **DoRA quality superior to LoRA especially in lower ranks** the difference in quality of DoRA of rank 8 and LoRA of rank 8 appears to be more significant than when training ranks of 32 or 64 for example.
+> This is also aligned with some of the quantitative analysis shown in the paper.
 
 **Usage**
-1. To use DoRA you need to install `peft` from main: 
+1. To use DoRA you need to install `peft` from main:
 ```bash
 pip install git+https://github.com/huggingface/peft.git
 ```
@@ -256,12 +256,12 @@ pip install git+https://github.com/huggingface/peft.git
 ```bash
 --use_dora
 ```
-**Inference** 
+**Inference**
 The inference is the same as if you train a regular LoRA 🤗
 
 ## Conducting EDM-style training
 
-It's now possible to perform EDM-style training as proposed in [Elucidating the Design Space of Diffusion-Based Generative Models](https://arxiv.org/abs/2206.00364). 
+It's now possible to perform EDM-style training as proposed in [Elucidating the Design Space of Diffusion-Based Generative Models](https://arxiv.org/abs/2206.00364).
 
 simply set:
 
@@ -304,12 +304,12 @@ accelerate launch train_dreambooth_lora_sdxl_advanced.py \
 > [!CAUTION]
 > Min-SNR gamma is not supported with the EDM-style training yet. When training with the PlaygroundAI model, it's recommended to not pass any "variant".
 
-### B-LoRA training 
+### B-LoRA training
 The advanced script now supports B-LoRA training too!
-> Proposed in [Implicit Style-Content Separation using B-LoRA](https://arxiv.org/abs/2403.14572), 
+> Proposed in [Implicit Style-Content Separation using B-LoRA](https://arxiv.org/abs/2403.14572),
 B-LoRA is a method that leverages LoRA to implicitly separate the style and content components of a **single** image.
-It was shown that learning the LoRA weights of two specific blocks (referred to as B-LoRAs) 
-achieves style-content separation that cannot be achieved by training each B-LoRA independently. 
+It was shown that learning the LoRA weights of two specific blocks (referred to as B-LoRAs)
+achieves style-content separation that cannot be achieved by training each B-LoRA independently.
 Once trained, the two B-LoRAs can be used as independent components to allow various image stylization tasks
 
 **Usage**
@@ -336,7 +336,7 @@ You can train a B-LoRA with as little as 1 image, and 1000 steps. Try this defau
  --gradient_checkpointing \
  --mixed_precision="fp16"
 ```
-**Inference** 
+**Inference**
 The inference is a bit different:
 1. we need load *specific* unet layers (as opposed to a regular LoRA/DoRA)
 2. the trained layers we load, changes based on our objective (e.g. style/content)
@@ -354,8 +354,8 @@ def is_belong_to_blocks(key, blocks):
         return False
     except Exception as e:
         raise type(e)(f'failed to is_belong_to_block, due to: {e}')
-    
-def lora_lora_unet_blocks(lora_path, alpha, target_blocks):  
+
+def lora_lora_unet_blocks(lora_path, alpha, target_blocks):
   state_dict, _ = pipeline.lora_state_dict(lora_path)
   filtered_state_dict = {k: v * alpha for k, v in state_dict.items() if is_belong_to_blocks(k, target_blocks)}
   return filtered_state_dict
@@ -367,7 +367,7 @@ pipeline = StableDiffusionXLPipeline.from_pretrained(
     torch_dtype=torch.float16,
 ).to("cuda")
 
-# pick a blora for content/style (you can also set one to None) 
+# pick a blora for content/style (you can also set one to None)
 content_B_lora_path  = "lora-library/B-LoRA-teddybear"
 style_B_lora_path= "lora-library/B-LoRA-pen_sketch"
 
@@ -384,28 +384,28 @@ prompt = "a [v18] in [v30] style"
 pipeline(prompt, num_images_per_prompt=4).images
 ```
 ### LoRA training of Targeted U-net Blocks
-The advanced script now supports custom choice of U-net blocks to train during Dreambooth LoRA tuning. 
+The advanced script now supports custom choice of U-net blocks to train during Dreambooth LoRA tuning.
 > [!NOTE]
 > This feature is still experimental
 
-> Recently, works like B-LoRA showed the potential advantages of learning the LoRA weights of specific U-net blocks, not only in speed & memory, 
-> but also in reducing the amount of needed data, improving style manipulation and overcoming overfitting issues. 
-> In light of this, we're introducing a new feature to the advanced script to allow for configurable U-net learned blocks. 
+> Recently, works like B-LoRA showed the potential advantages of learning the LoRA weights of specific U-net blocks, not only in speed & memory,
+> but also in reducing the amount of needed data, improving style manipulation and overcoming overfitting issues.
+> In light of this, we're introducing a new feature to the advanced script to allow for configurable U-net learned blocks.
 
 **Usage**
-Configure LoRA learned U-net blocks adding a `lora_unet_blocks` flag, with a comma seperated string specifying the targeted blocks. 
+Configure LoRA learned U-net blocks adding a `lora_unet_blocks` flag, with a comma seperated string specifying the targeted blocks.
 e.g:
 ```bash
 --lora_unet_blocks="unet.up_blocks.0.attentions.0,unet.up_blocks.0.attentions.1"
 ```
 
 > [!NOTE]
-> if you specify both `--use_blora` and `--lora_unet_blocks`, values given in --lora_unet_blocks will be ignored. 
-> When enabling --use_blora, targeted U-net blocks are automatically set to be "unet.up_blocks.0.attentions.0,unet.up_blocks.0.attentions.1" as discussed in the paper. 
+> if you specify both `--use_blora` and `--lora_unet_blocks`, values given in --lora_unet_blocks will be ignored.
+> When enabling --use_blora, targeted U-net blocks are automatically set to be "unet.up_blocks.0.attentions.0,unet.up_blocks.0.attentions.1" as discussed in the paper.
 > If you wish to experiment with different blocks, specify `--lora_unet_blocks` only.
 
-**Inference** 
-Inference is the same as for B-LoRAs, except the input targeted blocks should be modified based on your training configuration. 
+**Inference**
+Inference is the same as for B-LoRAs, except the input targeted blocks should be modified based on your training configuration.
 ```python
 import torch
 from diffusers import StableDiffusionXLPipeline, AutoencoderKL
@@ -419,8 +419,8 @@ def is_belong_to_blocks(key, blocks):
         return False
     except Exception as e:
         raise type(e)(f'failed to is_belong_to_block, due to: {e}')
-    
-def lora_lora_unet_blocks(lora_path, alpha, target_blocks):  
+
+def lora_lora_unet_blocks(lora_path, alpha, target_blocks):
   state_dict, _ = pipeline.lora_state_dict(lora_path)
   filtered_state_dict = {k: v * alpha for k, v in state_dict.items() if is_belong_to_blocks(k, target_blocks)}
   return filtered_state_dict
@@ -436,7 +436,7 @@ lora_path  = "lora-library/B-LoRA-pen_sketch"
 
 state_dict = lora_lora_unet_blocks(content_B_lora_path,alpha=1,target_blocks=["unet.up_blocks.0.attentions.0"])
 
-# Load traine dlora layers into the unet
+# Load trained lora layers into the unet
 pipeline.load_lora_into_unet(state_dict, None, pipeline.unet)
 
 #generate