Merge branch 'main' into patch-1

Author: a-r-r-o-w

.github/workflows/nightly_tests.yml

@@ -265,7 +265,7 @@ jobs:
 
       - name: Run PyTorch CUDA tests
         env:
-          HF_TOKEN: ${{ secrets.HF_TOKEN }}
+          HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_TOKEN }}
           # https://pytorch.org/docs/stable/notes/randomness.html#avoiding-nondeterministic-algorithms
           CUBLAS_WORKSPACE_CONFIG: :16:8
         run: |
@@ -505,7 +505,7 @@ jobs:
 #        shell: arch -arch arm64 bash {0}
 #        env:
 #          HF_HOME: /System/Volumes/Data/mnt/cache
-#          HF_TOKEN: ${{ secrets.HF_TOKEN }}
+#          HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_TOKEN }}
 #        run: |
 #          ${CONDA_RUN} python -m pytest -n 1 -s -v --make-reports=tests_torch_mps \
 #            --report-log=tests_torch_mps.log \
@@ -561,7 +561,7 @@ jobs:
 #        shell: arch -arch arm64 bash {0}
 #        env:
 #          HF_HOME: /System/Volumes/Data/mnt/cache
-#          HF_TOKEN: ${{ secrets.HF_TOKEN }}
+#          HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_TOKEN }}
 #        run: |
 #          ${CONDA_RUN} python -m pytest -n 1 -s -v --make-reports=tests_torch_mps \
 #            --report-log=tests_torch_mps.log \

.github/workflows/push_tests.yml

@@ -187,7 +187,7 @@ jobs:
 
     - name: Run Flax TPU tests
       env:
-        HF_TOKEN: ${{ secrets.HF_TOKEN }}
+        HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_TOKEN }}
       run: |
         python -m pytest -n 0 \
           -s -v -k "Flax" \
@@ -235,7 +235,7 @@ jobs:
 
     - name: Run ONNXRuntime CUDA tests
       env:
-        HF_TOKEN: ${{ secrets.HF_TOKEN }}
+        HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_TOKEN }}
       run: |
         python -m pytest -n 1 --max-worker-restart=0 --dist=loadfile \
           -s -v -k "Onnx" \
@@ -283,7 +283,7 @@ jobs:
         python utils/print_env.py
     - name: Run example tests on GPU
       env:
-        HF_TOKEN: ${{ secrets.HF_TOKEN }}
+        HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_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/
@@ -326,7 +326,7 @@ jobs:
         python utils/print_env.py
     - name: Run example tests on GPU
       env:
-        HF_TOKEN: ${{ secrets.HF_TOKEN }}
+        HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_TOKEN }}
       run: |
         python -m pytest -n 1 --max-worker-restart=0 --dist=loadfile -s -v -k "xformers" --make-reports=tests_torch_xformers_cuda tests/
     - name: Failure short reports
@@ -372,7 +372,7 @@ jobs:
 
     - name: Run example tests on GPU
       env:
-        HF_TOKEN: ${{ secrets.HF_TOKEN }}
+        HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_TOKEN }}
       run: |
         python -m venv /opt/venv && export PATH="/opt/venv/bin:$PATH"
         python -m uv pip install timm

.github/workflows/release_tests_fast.yml

@@ -81,7 +81,7 @@ jobs:
           python utils/print_env.py
       - name: Slow PyTorch CUDA checkpoint tests on Ubuntu
         env:
-          HF_TOKEN: ${{ secrets.HF_TOKEN }}
+          HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_TOKEN }}
           # https://pytorch.org/docs/stable/notes/randomness.html#avoiding-nondeterministic-algorithms
           CUBLAS_WORKSPACE_CONFIG: :16:8
         run: |
@@ -135,7 +135,7 @@ jobs:
 
     - name: Run PyTorch CUDA tests
       env:
-        HF_TOKEN: ${{ secrets.HF_TOKEN }}
+        HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_TOKEN }}
         # https://pytorch.org/docs/stable/notes/randomness.html#avoiding-nondeterministic-algorithms
         CUBLAS_WORKSPACE_CONFIG: :16:8
       run: |
@@ -186,7 +186,7 @@ jobs:
 
       - name: Run PyTorch CUDA tests
         env:
-          HF_TOKEN: ${{ secrets.HF_TOKEN }}
+          HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_TOKEN }}
           # https://pytorch.org/docs/stable/notes/randomness.html#avoiding-nondeterministic-algorithms
           CUBLAS_WORKSPACE_CONFIG: :16:8
         run: |
@@ -241,7 +241,7 @@ jobs:
 
     - name: Run slow Flax TPU tests
       env:
-        HF_TOKEN: ${{ secrets.HF_TOKEN }}
+        HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_TOKEN }}
       run: |
         python -m pytest -n 0 \
           -s -v -k "Flax" \
@@ -289,7 +289,7 @@ jobs:
 
     - name: Run slow ONNXRuntime CUDA tests
       env:
-        HF_TOKEN: ${{ secrets.HF_TOKEN }}
+        HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_TOKEN }}
       run: |
         python -m pytest -n 1 --max-worker-restart=0 --dist=loadfile \
           -s -v -k "Onnx" \
@@ -337,7 +337,7 @@ jobs:
         python utils/print_env.py
     - name: Run example tests on GPU
       env:
-        HF_TOKEN: ${{ secrets.HF_TOKEN }}
+        HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_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/
@@ -380,7 +380,7 @@ jobs:
         python utils/print_env.py
     - name: Run example tests on GPU
       env:
-        HF_TOKEN: ${{ secrets.HF_TOKEN }}
+        HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_TOKEN }}
       run: |
         python -m pytest -n 1 --max-worker-restart=0 --dist=loadfile -s -v -k "xformers" --make-reports=tests_torch_xformers_cuda tests/
     - name: Failure short reports
@@ -426,7 +426,7 @@ jobs:
 
     - name: Run example tests on GPU
       env:
-        HF_TOKEN: ${{ secrets.HF_TOKEN }}
+        HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_TOKEN }}
       run: |
         python -m venv /opt/venv && export PATH="/opt/venv/bin:$PATH"
         python -m uv pip install timm

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

@@ -309,6 +309,53 @@ image.save("output.png")
 
 When unloading the Control LoRA weights, call `pipe.unload_lora_weights(reset_to_overwritten_params=True)` to reset the `pipe.transformer` completely back to its original form. The resultant pipeline can then be used with methods like [`DiffusionPipeline.from_pipe`]. More details about this argument are available in [this PR](https://github.com/huggingface/diffusers/pull/10397).
 
+## IP-Adapter
+
+<Tip>
+
+Check out [IP-Adapter](../../../using-diffusers/ip_adapter) to learn more about how IP-Adapters work.
+
+</Tip>
+
+An IP-Adapter lets you prompt Flux with images, in addition to the text prompt. This is especially useful when describing complex concepts that are difficult to articulate through text alone and you have reference images.
+
+```python
+import torch
+from diffusers import FluxPipeline
+from diffusers.utils import load_image
+
+pipe = FluxPipeline.from_pretrained(
+    "black-forest-labs/FLUX.1-dev", torch_dtype=torch.bfloat16
+).to("cuda")
+
+image = load_image("https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/flux_ip_adapter_input.jpg").resize((1024, 1024))
+
+pipe.load_ip_adapter(
+    "XLabs-AI/flux-ip-adapter",
+    weight_name="ip_adapter.safetensors",
+    image_encoder_pretrained_model_name_or_path="openai/clip-vit-large-patch14"
+)
+pipe.set_ip_adapter_scale(1.0)
+
+image = pipe(
+    width=1024,
+    height=1024,
+    prompt="wearing sunglasses",
+    negative_prompt="",
+    true_cfg=4.0,
+    generator=torch.Generator().manual_seed(4444),
+    ip_adapter_image=image,
+).images[0]
+
+image.save('flux_ip_adapter_output.jpg')
+```
+
+<div class="justify-center">
+    <img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/flux_ip_adapter_output.jpg"/>
+    <figcaption class="mt-2 text-sm text-center text-gray-500">IP-Adapter examples with prompt "wearing sunglasses"</figcaption>
+</div>
+
+
 ## Running FP16 inference
 
 Flux can generate high-quality images with FP16 (i.e. to accelerate inference on Turing/Volta GPUs) but produces different outputs compared to FP32/BF16. The issue is that some activations in the text encoders have to be clipped when running in FP16, which affects the overall image. Forcing text encoders to run with FP32 inference thus removes this output difference. See [here](https://github.com/huggingface/diffusers/pull/9097#issuecomment-2272292516) for details.

docs/source/en/installation.md

@@ -23,32 +23,60 @@ You should install 🤗 Diffusers in a [virtual environment](https://docs.python
 If you're unfamiliar with Python virtual environments, take a look at this [guide](https://packaging.python.org/guides/installing-using-pip-and-virtual-environments/).
 A virtual environment makes it easier to manage different projects and avoid compatibility issues between dependencies.
 
-Start by creating a virtual environment in your project directory:
+Create a virtual environment with Python or [uv](https://docs.astral.sh/uv/) (refer to [Installation](https://docs.astral.sh/uv/getting-started/installation/) for installation instructions), a fast Rust-based Python package and project manager.
+
+<hfoptions id="install">
+<hfoption id="uv">
 
 ```bash
-python -m venv .env
+uv venv my-env
+source my-env/bin/activate
 ```
 
-Activate the virtual environment:
+</hfoption>
+<hfoption id="Python">
 
 ```bash
-source .env/bin/activate
+python -m venv my-env
+source my-env/bin/activate
 ```
 
-You should also install 🤗 Transformers because 🤗 Diffusers relies on its models:
+</hfoption>
+</hfoptions>
+
+You should also install 🤗 Transformers because 🤗 Diffusers relies on its models.
 
 
 <frameworkcontent>
 <pt>
-Note - PyTorch only supports Python 3.8 - 3.11 on Windows.
+
+PyTorch only supports Python 3.8 - 3.11 on Windows. Install Diffusers with uv.
+
+```bash
+uv install diffusers["torch"] transformers
+```
+
+You can also install Diffusers with pip.
+
 ```bash
 pip install diffusers["torch"] transformers
 ```
+
 </pt>
 <jax>
+
+Install Diffusers with uv.
+
+```bash
+uv pip install diffusers["flax"] transformers
+```
+
+You can also install Diffusers with pip.
+
 ```bash
 pip install diffusers["flax"] transformers
 ```
+
 </jax>
 </frameworkcontent>
 

docs/source/en/optimization/para_attn.md

@@ -29,7 +29,7 @@ However, it is hard to decide when to reuse the cache to ensure quality generate
 This achieves a 2x speedup on FLUX.1-dev and HunyuanVideo inference with very good quality.
 
 <figure>
-    <img src="https://huggingface.co/datasets/chengzeyi/documentation-images/resolve/main/diffusers/para-attn/ada-cache.png" alt="Cache in Diffusion Transformer" />
+    <img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/para-attn/ada-cache.png" alt="Cache in Diffusion Transformer" />
     <figcaption>How AdaCache works, First Block Cache is a variant of it</figcaption>
 </figure>
 

src/diffusers/pipelines/flux/pipeline_flux_controlnet_inpainting.py

@@ -930,8 +930,8 @@ def __call__(
         if isinstance(self.controlnet, FluxControlNetModel):
             control_image = self.prepare_image(
                 image=control_image,
-                width=height,
-                height=width,
+                width=width,
+                height=height,
                 batch_size=batch_size * num_images_per_prompt,
                 num_images_per_prompt=num_images_per_prompt,
                 device=device,

src/diffusers/utils/import_utils.py

@@ -149,6 +149,7 @@
         "onnxruntime-openvino",
         "ort_nightly_directml",
         "onnxruntime-rocm",
+        "onnxruntime-migraphx",
         "onnxruntime-training",
     )
     _onnxruntime_version = None

tests/schedulers/test_scheduler_flax.py

@@ -338,8 +338,8 @@ def test_full_loop_no_noise(self):
             assert abs(result_sum - 255.0714) < 1e-2
             assert abs(result_mean - 0.332124) < 1e-3
         else:
-            assert abs(result_sum - 255.1113) < 1e-1
-            assert abs(result_mean - 0.332176) < 1e-3
+            assert abs(result_sum - 270.2) < 1e-1
+            assert abs(result_mean - 0.3519494) < 1e-3
 
 
 @require_flax