Merge branch 'main' into main

Author: hlky

.github/workflows/nightly_tests.yml

@@ -359,6 +359,8 @@ jobs:
             test_location: "bnb"
           - backend: "gguf"
             test_location: "gguf"
+          - backend: "torchao"
+            test_location: "torchao"
     runs-on:
       group: aws-g6e-xlarge-plus
     container:

.github/workflows/pypi_publish.yaml

@@ -68,7 +68,7 @@ jobs:
       - name: Test installing diffusers and importing
         run: |
           pip install diffusers && pip uninstall diffusers -y
-          pip install -i https://testpypi.python.org/pypi diffusers
+          pip install -i https://test.pypi.org/simple/ diffusers
           python -c "from diffusers import __version__; print(__version__)"
           python -c "from diffusers import DiffusionPipeline; pipe = DiffusionPipeline.from_pretrained('fusing/unet-ldm-dummy-update'); pipe()"
           python -c "from diffusers import DiffusionPipeline; pipe = DiffusionPipeline.from_pretrained('hf-internal-testing/tiny-stable-diffusion-pipe', safety_checker=None); pipe('ah suh du')"

docs/source/en/_toctree.yml

@@ -429,7 +429,7 @@
     - local: api/pipelines/ledits_pp
       title: LEDITS++
     - local: api/pipelines/ltx_video
-      title: LTX
+      title: LTXVideo
     - local: api/pipelines/lumina
       title: Lumina-T2X
     - local: api/pipelines/marigold

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

@@ -18,7 +18,7 @@ The model can be loaded with the following code snippet.
 ```python
 from diffusers import AutoencoderKLLTXVideo
 
-vae = AutoencoderKLLTXVideo.from_pretrained("TODO/TODO", subfolder="vae", torch_dtype=torch.float32).to("cuda")
+vae = AutoencoderKLLTXVideo.from_pretrained("Lightricks/LTX-Video", subfolder="vae", torch_dtype=torch.float32).to("cuda")
 ```
 
 ## AutoencoderKLLTXVideo

docs/source/en/api/models/ltx_video_transformer3d.md

@@ -18,7 +18,7 @@ The model can be loaded with the following code snippet.
 ```python
 from diffusers import LTXVideoTransformer3DModel
 
-transformer = LTXVideoTransformer3DModel.from_pretrained("TODO/TODO", subfolder="transformer", torch_dtype=torch.bfloat16).to("cuda")
+transformer = LTXVideoTransformer3DModel.from_pretrained("Lightricks/LTX-Video", subfolder="transformer", torch_dtype=torch.bfloat16).to("cuda")
 ```
 
 ## LTXVideoTransformer3DModel

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

@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License. -->
 
-# LTX
+# LTX Video
 
 [LTX Video](https://huggingface.co/Lightricks/LTX-Video) is the first DiT-based video generation model capable of generating high-quality videos in real-time. It produces 24 FPS videos at a 768x512 resolution faster than they can be watched. Trained on a large-scale dataset of diverse videos, the model generates high-resolution videos with realistic and varied content. We provide a model for both text-to-video as well as image + text-to-video usecases.
 
@@ -22,14 +22,24 @@ Make sure to check out the Schedulers [guide](../../using-diffusers/schedulers.m
 
 </Tip>
 
+Available models:
+
+|  Model name   | Recommended dtype |
+|:-------------:|:-----------------:|
+| [`LTX Video 0.9.0`](https://huggingface.co/Lightricks/LTX-Video/blob/main/ltx-video-2b-v0.9.safetensors) | `torch.bfloat16` |
+| [`LTX Video 0.9.1`](https://huggingface.co/Lightricks/LTX-Video/blob/main/ltx-video-2b-v0.9.1.safetensors) | `torch.bfloat16` |
+
+Note: The recommended dtype is for the transformer component. The VAE and text encoders can be either `torch.float32`, `torch.bfloat16` or `torch.float16` but the recommended dtype is `torch.bfloat16` as used in the original repository.
+
 ## Loading Single Files
 
-Loading the original LTX Video checkpoints is also possible with [`~ModelMixin.from_single_file`].
+Loading the original LTX Video checkpoints is also possible with [`~ModelMixin.from_single_file`]. We recommend using `from_single_file` for the Lightricks series of models, as they plan to release multiple models in the future in the single file format.
 
 ```python
 import torch
 from diffusers import AutoencoderKLLTXVideo, LTXImageToVideoPipeline, LTXVideoTransformer3DModel
 
+# `single_file_url` could also be https://huggingface.co/Lightricks/LTX-Video/ltx-video-2b-v0.9.1.safetensors
 single_file_url = "https://huggingface.co/Lightricks/LTX-Video/ltx-video-2b-v0.9.safetensors"
 transformer = LTXVideoTransformer3DModel.from_single_file(
   single_file_url, torch_dtype=torch.bfloat16
@@ -99,6 +109,34 @@ export_to_video(video, "output_gguf_ltx.mp4", fps=24)
 
 Make sure to read the [documentation on GGUF](../../quantization/gguf) to learn more about our GGUF support.
 
+<!-- TODO(aryan): Update this when official weights are supported -->
+
+Loading and running inference with [LTX Video 0.9.1](https://huggingface.co/Lightricks/LTX-Video/blob/main/ltx-video-2b-v0.9.1.safetensors) weights.
+
+```python
+import torch
+from diffusers import LTXPipeline
+from diffusers.utils import export_to_video
+
+pipe = LTXPipeline.from_pretrained("a-r-r-o-w/LTX-Video-0.9.1-diffusers", torch_dtype=torch.bfloat16)
+pipe.to("cuda")
+
+prompt = "A woman with long brown hair and light skin smiles at another woman with long blonde hair. The woman with brown hair wears a black jacket and has a small, barely noticeable mole on her right cheek. The camera angle is a close-up, focused on the woman with brown hair's face. The lighting is warm and natural, likely from the setting sun, casting a soft glow on the scene. The scene appears to be real-life footage"
+negative_prompt = "worst quality, inconsistent motion, blurry, jittery, distorted"
+
+video = pipe(
+    prompt=prompt,
+    negative_prompt=negative_prompt,
+    width=768,
+    height=512,
+    num_frames=161,
+    decode_timestep=0.03,
+    decode_noise_scale=0.025,
+    num_inference_steps=50,
+).frames[0]
+export_to_video(video, "output.mp4", fps=24)
+```
+
 Refer to [this section](https://huggingface.co/docs/diffusers/main/en/api/pipelines/cogvideox#memory-optimization) to learn more about optimizing memory consumption.
 
 ## LTXPipeline

docs/source/en/community_projects.md

@@ -79,4 +79,8 @@ Happy exploring, and thank you for being part of the Diffusers community!
     <td><a href="https://github.com/Netwrck/stable-diffusion-server"> Stable Diffusion Server </a></td>
     <td>A server configured for Inpainting/Generation/img2img with one stable diffusion model</td>
   </tr>
+  <tr style="border-top: 2px solid black">
+    <td><a href="https://github.com/suzukimain/auto_diffusers"> Model Search </a></td>
+    <td>Search models on Civitai and Hugging Face</td>
+  </tr>
 </table>

docs/source/en/quantization/torchao.md

@@ -25,6 +25,7 @@ Quantize a model by passing [`TorchAoConfig`] to [`~ModelMixin.from_pretrained`]
 The example below only quantizes the weights to int8.
 
 ```python
+import torch
 from diffusers import FluxPipeline, FluxTransformer2DModel, TorchAoConfig
 
 model_id = "black-forest-labs/FLUX.1-dev"
@@ -44,6 +45,10 @@ pipe = FluxPipeline.from_pretrained(
 )
 pipe.to("cuda")
 
+# Without quantization: ~31.447 GB
+# With quantization: ~20.40 GB
+print(f"Pipeline memory usage: {torch.cuda.max_memory_reserved() / 1024**3:.3f} GB")
+
 prompt = "A cat holding a sign that says hello world"
 image = pipe(
     prompt, num_inference_steps=50, guidance_scale=4.5, max_sequence_length=512
@@ -88,6 +93,63 @@ Some quantization methods are aliases (for example, `int8wo` is the commonly use
 
 Refer to the official torchao documentation for a better understanding of the available quantization methods and the exhaustive list of configuration options available.
 
+## Serializing and Deserializing quantized models
+
+To serialize a quantized model in a given dtype, first load the model with the desired quantization dtype and then save it using the [`~ModelMixin.save_pretrained`] method.
+
+```python
+import torch
+from diffusers import FluxTransformer2DModel, TorchAoConfig
+
+quantization_config = TorchAoConfig("int8wo")
+transformer = FluxTransformer2DModel.from_pretrained(
+    "black-forest-labs/Flux.1-Dev",
+    subfolder="transformer",
+    quantization_config=quantization_config,
+    torch_dtype=torch.bfloat16,
+)
+transformer.save_pretrained("/path/to/flux_int8wo", safe_serialization=False)
+```
+
+To load a serialized quantized model, use the [`~ModelMixin.from_pretrained`] method.
+
+```python
+import torch
+from diffusers import FluxPipeline, FluxTransformer2DModel
+
+transformer = FluxTransformer2DModel.from_pretrained("/path/to/flux_int8wo", torch_dtype=torch.bfloat16, use_safetensors=False)
+pipe = FluxPipeline.from_pretrained("black-forest-labs/Flux.1-Dev", transformer=transformer, torch_dtype=torch.bfloat16)
+pipe.to("cuda")
+
+prompt = "A cat holding a sign that says hello world"
+image = pipe(prompt, num_inference_steps=30, guidance_scale=7.0).images[0]
+image.save("output.png")
+```
+
+Some quantization methods, such as `uint4wo`, cannot be loaded directly and may result in an `UnpicklingError` when trying to load the models, but work as expected when saving them. In order to work around this, one can load the state dict manually into the model. Note, however, that this requires using `weights_only=False` in `torch.load`, so it should be run only if the weights were obtained from a trustable source.
+
+```python
+import torch
+from accelerate import init_empty_weights
+from diffusers import FluxPipeline, FluxTransformer2DModel, TorchAoConfig
+
+# Serialize the model
+transformer = FluxTransformer2DModel.from_pretrained(
+    "black-forest-labs/Flux.1-Dev",
+    subfolder="transformer",
+    quantization_config=TorchAoConfig("uint4wo"),
+    torch_dtype=torch.bfloat16,
+)
+transformer.save_pretrained("/path/to/flux_uint4wo", safe_serialization=False, max_shard_size="50GB")
+# ...
+
+# Load the model
+state_dict = torch.load("/path/to/flux_uint4wo/diffusion_pytorch_model.bin", weights_only=False, map_location="cpu")
+with init_empty_weights():
+    transformer = FluxTransformer2DModel.from_config("/path/to/flux_uint4wo/config.json")
+transformer.load_state_dict(state_dict, strict=True, assign=True)
+```
+
 ## Resources
 
 - [TorchAO Quantization API](https://github.com/pytorch/ao/blob/main/torchao/quantization/README.md)

examples/advanced_diffusion_training/train_dreambooth_lora_flux_advanced.py

@@ -74,7 +74,7 @@
     import wandb
 
 # Will error if the minimal version of diffusers is not installed. Remove at your own risks.
-check_min_version("0.32.0.dev0")
+check_min_version("0.33.0.dev0")
 
 logger = get_logger(__name__)
 

examples/advanced_diffusion_training/train_dreambooth_lora_sd15_advanced.py

@@ -73,7 +73,7 @@
 
 
 # Will error if the minimal version of diffusers is not installed. Remove at your own risks.
-check_min_version("0.32.0.dev0")
+check_min_version("0.33.0.dev0")
 
 logger = get_logger(__name__)