Skip to content

Releases: huggingface/diffusers

v0.4.1: Patch release

07 Oct 09:01
@patrickvonplaten patrickvonplaten
v0.4.1
91ddd2a
Compare
Choose a tag to compare
v0.4.1: Patch release

This patch release fixes an bug with incorrect module naming for community pipelines and an incorrect breaking change when moving piplines in fp16 to "cpu" or "mps".

Contributors

apolinario and patrickvonplaten
Assets 2
Loading

v0.4.0 Better, faster, stronger!

06 Oct 16:37
@patrickvonplaten patrickvonplaten
v0.4.0
0fe59b6
Compare
Choose a tag to compare
v0.4.0 Better, faster, stronger!

🚗 Faster

We have thoroughly profiled our codebase and applied a number of incremental improvements that, when combined, provide a speed improvement of almost 3x.

On top of that, we now default to using the float16 format. It's much faster than float32 and, according to our tests, produces images with no discernible difference in quality. This beats the use of autocast, so the resulting code is cleaner!

🔑 use_auth_token no more

The recently released version of huggingface-hub automatically uses your access token if you are logged in, so you don't need to put it everywhere in your code. All you need to do is authenticate once using huggingface-cli login in your terminal and you're all set.

- pipe = DiffusionPipeline.from_pretrained("CompVis/stable-diffusion-v1-4", use_auth_token=True)
+ pipe = DiffusionPipeline.from_pretrained("CompVis/stable-diffusion-v1-4")

We bumped huggingface-hub version to 0.10.0 in our dependencies to achieve this.

🎈More flexible APIs

  • Schedulers now use a common, simpler unified API design. This has allowed us to remove many conditionals and special cases in the rest of the code, including the pipelines. This is very important for us and for the users of 🧨 diffusers: we all gain clarity and a solid abstraction for schedulers. See the description in #719 for more details

Please update any custom Stable Diffusion pipelines accordingly:

- if isinstance(self.scheduler, LMSDiscreteScheduler):
-    latents = latents * self.scheduler.sigmas[0]
+ latents = latents * self.scheduler.init_noise_sigma
- if isinstance(self.scheduler, LMSDiscreteScheduler):
-     sigma = self.scheduler.sigmas[i]
-     latent_model_input = latent_model_input / ((sigma**2 + 1) ** 0.5)
+ latent_model_input = self.scheduler.scale_model_input(latent_model_input, t)
- if isinstance(self.scheduler, LMSDiscreteScheduler):
-     latents = self.scheduler.step(noise_pred, i, latents, **extra_step_kwargs).prev_sample
- else:
-     latents = self.scheduler.step(noise_pred, t, latents, **extra_step_kwargs).prev_sample
+ latents = self.scheduler.step(noise_pred, t, latents, **extra_step_kwargs).prev_sample
  • Pipeline callbacks. As a community project (h/t @jamestiotio!), diffusers pipelines can now invoke a callback function during generation, providing the latents at each step of the process. This makes it easier to perform tasks such as visualization, inspection, explainability and others the community may invent.

🛠️ More tasks

Building on top of the previous foundations, this release incorporates several new tasks that have been adapted from research papers or community projects. These include:

  • Textual inversion. Makes it possible to quickly train a new concept or style and incorporate it into the vocabulary of Stable Diffusion. Hundreds of people have already created theirs, and they can be shared and combined together. See the training Colab to get started.
  • Dreambooth. Similar goal to textual inversion, but instead of creating a new item in the vocabulary it fine-tunes the model to make it learn a new concept. Training Colab.
  • Negative prompts. Another community effort led by @shirayu. The Stable Diffusion pipeline can now receive both a positive prompt (the one you want to create), and a negative prompt (something you want to drive the model away from). This opens up a lot of creative possibilities!

🏃‍♀️ Under the hood changes to support better fine-tuning

Gradient checkpointing and 8-bit optimizers have been successfully applied to achieve Dreambooth fine-tuning in a Colab notebook! These updates will make it easier for diffusers to support general-purpose fine-tuning (coming soon!).

⚠️ Experimental: community pipelines

This is big, but it's still an experimental feature that may change in the future.

We are constantly amazed at the amount of imagination and creativity in the diffusers community, so we've made it easy to create custom pipelines and share them with others. You can write your own pipeline code, store it in 🤗 Hub, GitHub or your local filesystem and StableDiffusionPipeline.from_pretrained will be able to load and run it. Read more in the documentation.

We can't wait to see what new tasks the community creates!

💪 Quality of life fixes

Bug fixing, improved documentation, better tests are all important to ensure diffusers is a high-quality codebase, and we always spend a lot of effort working on them. Several first-time contributors have helped here, and we are very grateful for their efforts!

🙏 Significant community contributions

The following people have made significant contributions to the library over the last release:

  • @Victarry – Add training example for DreamBooth (#554)
  • @jamestiotio – Add callback parameters for Stable Diffusion pipelines (#521)
  • @jachiam – Allow resolutions that are not multiples of 64 (#505)
  • @johnowhitaker – Adding pred_original_sample to SchedulerOutput for some samplers (#614).
  • @keturn – Interesting discussions and insights on many topics.

✏️ Change list

Read more

Contributors

kashif, keturn, and 34 other contributors
Assets 2
Loading

v0.3.0: New API, Stable Diffusion pipelines, low-memory inference, MPS backend, ONNX

08 Sep 17:09
@anton-l anton-l
v0.3.0
3f55d13
Compare
Choose a tag to compare
v0.3.0: New API, Stable Diffusion pipelines, low-memory inference, MPS backend, ONNX

📚 Shiny new docs!

Thanks to the community efforts for [Docs] and [Type Hints] we've started populating the Diffusers documentation pages with lots of helpful guides, links and API references.

📝 New API & breaking changes

New API

Pipeline, Model, and Scheduler outputs can now be both dataclasses, Dicts, and Tuples:

image = pipe("The red cat is sitting on a chair")["sample"][0]

is now replaced by:

image = pipe("The red cat is sitting on a chair").images[0]
# or
image = pipe("The red cat is sitting on a chair")["image"][0]
# or
image = pipe("The red cat is sitting on a chair")[0]

Similarly:

sample = unet(...).sample

and

prev_sample = scheduler(...).prev_sample

is now possible!

🚨🚨🚨 Breaking change 🚨🚨🚨

This PR introduces breaking changes for the following public-facing methods:

  • VQModel.encode -> we return a dict/dataclass instead of a single tensor. In the future it's very likely required to return more than just one tensor. Please make sure to change latents = model.encode(...) to latents = model.encode(...)[0] or latents = model.encode(...).latens
  • VQModel.decode -> we return a dict/dataclass instead of a single tensor. In the future it's very likely required to return more than just one tensor. Please make sure to change sample = model.decode(...) to sample = model.decode(...)[0] or sample = model.decode(...).sample
  • VQModel.forward -> we return a dict/dataclass instead of a single tensor. In the future it's very likely required to return more than just one tensor. Please make sure to change sample = model(...) to sample = model(...)[0] or sample = model(...).sample
  • AutoencoderKL.encode -> we return a dict/dataclass instead of a single tensor. In the future it's very likely required to return more than just one tensor. Please make sure to change latent_dist = model.encode(...) to latent_dist = model.encode(...)[0] or latent_dist = model.encode(...).latent_dist
  • AutoencoderKL.decode -> we return a dict/dataclass instead of a single tensor. In the future it's very likely required to return more than just one tensor. Please make sure to change sample = model.decode(...) to sample = model.decode(...)[0] or sample = model.decode(...).sample
  • AutoencoderKL.forward -> we return a dict/dataclass instead of a single tensor. In the future it's very likely required to return more than just one tensor. Please make sure to change sample = model(...) to sample = model(...)[0] or sample = model(...).sample

🎨 New Stable Diffusion pipelines

A couple of new pipelines have been added to Diffusers! We invite you to experiment with them, and to take them as inspiration to create your cool new tasks. These are the new pipelines:

  • Image-to-image generation. In addition to using a text prompt, this pipeline lets you include an example image to be used as the initial state of the process. 🤗 Diffuse the Rest is a cool demo about it!
  • Inpainting (experimental). You can provide an image and a mask and ask Stable Diffusion to replace the mask.

For more details about how they work, please visit our new API documentation.

This is a summary of all the Stable Diffusion tasks that can be easily used with 🤗 Diffusers:

Pipeline Tasks Colab Demo
pipeline_stable_diffusion.py Text-to-Image Generation Open In Colab 🤗 Stable Diffusion
pipeline_stable_diffusion_img2img.py Image-to-Image Text-Guided Generation Open In Colab 🤗 Diffuse the Rest
pipeline_stable_diffusion_inpaint.py Experimental – Text-Guided Image Inpainting Open In Colab Coming soon

🍬 Less memory usage for smaller GPUs

Now the diffusion models can take up significantly less VRAM (3.2 GB for Stable Diffusion) at the expense of 10% of speed thanks to the optimizations discussed in basujindal/stable-diffusion#117.

To make use of the attention optimization, just enable it with .enable_attention_slicing() after loading the pipeline:

from diffusers import StableDiffusionPipeline

pipe = StableDiffusionPipeline.from_pretrained(
    "CompVis/stable-diffusion-v1-4", 
    revision="fp16", 
    torch_dtype=torch.float16,
    use_auth_token=True
)
pipe = pipe.to("cuda")
pipe.enable_attention_slicing()

This will allow many more users to play with Stable Diffusion in their own computers! We can't wait to see what new ideas and results will be created by the community!

🐈‍⬛ Textual Inversion

Textual Inversion lets you personalize a Stable Diffusion model on your own images with just 3-5 samples.

GitHub: https://github.com/huggingface/diffusers/tree/main/examples/textual_inversion
Training: https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers/sd_textual_inversion_training.ipynb
Inference: https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers/stable_conceptualizer_inference.ipynb

🍎 MPS backend for Apple Silicon

🤗 Diffusers is compatible with Apple silicon for Stable Diffusion inference, using the PyTorch mps device. You need to install PyTorch Preview (Nightly) on a Mac with M1 or M2 CPU, and then use the pipeline as usual:

from diffusers import StableDiffusionPipeline

pipe = StableDiffusionPipeline.from_pretrained("CompVis/stable-diffusion-v1-4", use_auth_token=True)
pipe = pipe.to("mps")

prompt = "a photo of an astronaut riding a horse on mars"
image = pipe(prompt).images[0]

We are seeing great speedups (31s vs 214s in a M1 Max), but there are still a couple of limitations. We encourage you to read the documentation for the details.

🏭 Experimental ONNX exporter and pipeline for Stable Diffusion

We introduce a new (and experimental) Stable Diffusion pipeline compatible with the ONNX Runtime. This allows you to run Stable Diffusion on any hardware that supports ONNX (including a significant speedup on CPUs).

You need to use StableDiffusionOnnxPipeline instead of StableDiffusionPipeline. You also need to download the weights from the onnx branch of the repository, and indicate the runtime provider you want to use (CPU, in the following example):

from diffusers import StableDiffusionOnnxPipeline

pipe = StableDiffusionOnnxPipeline.from_pretrained(
    "CompVis/stable-diffusion-v1-4",
    revision="onnx",
    provider="CPUExecutionProvider",
    use_auth_token=True,
)

prompt = "a photo of an astronaut riding a horse on mars"
image = pipe(prompt).images[0]

⚠️ Warning: the script above takes a long time to download the external ONNX weights, so it will be faster to convert the checkpoint yourself (see below).

To convert your own checkpoint, run the conversion script locally:

python scripts/convert_stable_diffusion_checkpoint_to_onnx.py --model_path="CompVis/stable-diffusion-v1-4" --output_path="./stable_diffusion_onnx"

After that it can be loaded from the local path:

pipe = StableDiffusionOnnxPipeline.from_pretrained("./stable_diffusion_onnx", provider="CPUExecutionProvider")

Improvements and bugfixes

Read more

Contributors

kashif, apolinario, and 26 other contributors
Assets 2
Loading

v0.2.4: Patch release

22 Aug 17:09
@anton-l anton-l
v0.2.4
60a1473
Compare
Choose a tag to compare
v0.2.4: Patch release

This patch release allows the Stable Diffusion pipelines to be loaded with float16 precision:

pipe = StableDiffusionPipeline.from_pretrained(
           "CompVis/stable-diffusion-v1-4", 
           revision="fp16", 
           torch_dtype=torch.float16, 
           use_auth_token=True
)
pipe = pipe.to("cuda")

The resulting models take up less than 6900 MiB of GPU memory.

Contributors

patrickvonplaten
Assets 2
Loading

v0.2.3: Stable Diffusion public release

22 Aug 08:59
@anton-l anton-l
v0.2.3
ebd80e2
Compare
Choose a tag to compare
v0.2.3: Stable Diffusion public release

🎨 Stable Diffusion public release

The Stable Diffusion checkpoints are now public and can be loaded by anyone! 🥳

Make sure to accept the license terms on the model page first (requires login): https://huggingface.co/CompVis/stable-diffusion-v1-4
Install the required packages: pip install diffusers==0.2.3 transformers scipy
And log in on your machine using the huggingface-cli login command.

from torch import autocast
from diffusers import StableDiffusionPipeline, LMSDiscreteScheduler

# this will substitute the default PNDM scheduler for K-LMS  
lms = LMSDiscreteScheduler(
    beta_start=0.00085, 
    beta_end=0.012, 
    beta_schedule="scaled_linear"
)

pipe = StableDiffusionPipeline.from_pretrained(
    "CompVis/stable-diffusion-v1-4", 
    scheduler=lms,
    use_auth_token=True
).to("cuda")

prompt = "a photo of an astronaut riding a horse on mars"
with autocast("cuda"):
    image = pipe(prompt)["sample"][0]  
    
image.save("astronaut_rides_horse.png")

The safety checker

Following the model authors' guidelines and code, the Stable Diffusion inference results will now be filtered to exclude unsafe content. Any images classified as unsafe will be returned as blank. To check if the safety module is triggered programmaticaly, check the nsfw_content_detected flag like so:

outputs = pipe(prompt)
image = outputs
if any(outputs["nsfw_content_detected"]):
    print("Potential unsafe content was detected in one or more images. Try again with a different prompt and/or seed.")

Improvements and bugfixes

Full Changelog: v0.2.2...v0.2.3

Contributors

apolinario, pcuenca, and 3 other contributors
Assets 2
Loading

v0.2.2

16 Aug 17:59
@patrickvonplaten patrickvonplaten
v0.2.2
4b02f53
Compare
Choose a tag to compare
v0.2.2

This patch release fixes an import of the StableDiffusionPipeline

[K-LMS Scheduler] fix import by @patrickvonplaten in #191

Contributors

patrickvonplaten
Assets 2
Loading

v0.2.1 Patch release

16 Aug 16:32
@patrickvonplaten patrickvonplaten
v0.2.1
c1b378d
Compare
Choose a tag to compare
v0.2.1 Patch release

This patch release fixes a small bug of the StableDiffusionPipeline

Contributors

patrickvonplaten
Assets 2
Loading

v0.2.0: Stable Diffusion early access, K-LMS sampling

16 Aug 15:52
@anton-l anton-l
v0.2.0
ea2e177
Compare
Choose a tag to compare
v0.2.0: Stable Diffusion early access, K-LMS sampling

Stable Diffusion

Stable Diffusion is a text-to-image latent diffusion model created by the researchers and engineers from CompVis, Stability AI and LAION. It's trained on 512x512 images from a subset of the LAION-5B database. This model uses a frozen CLIP ViT-L/14 text encoder to condition the model on text prompts. With its 860M UNet and 123M text encoder, the model is relatively lightweight and runs on a GPU with at least 10GB VRAM.
See the model card for more information.

The Stable Diffusion weights are currently only available to universities, academics, research institutions and independent researchers. Please request access applying to this form

from torch import autocast
from diffusers import StableDiffusionPipeline

# make sure you're logged in with `huggingface-cli login`
pipe = StableDiffusionPipeline.from_pretrained("CompVis/stable-diffusion-v1-3-diffusers", use_auth_token=True)  

prompt = "a photograph of an astronaut riding a horse"
with autocast("cuda"):
    image = pipe(prompt, guidance_scale=7)["sample"][0]  # image here is in PIL format
    
image.save(f"astronaut_rides_horse.png")

K-LMS sampling

The new LMSDiscreteScheduler is a port of k-lms from k-diffusion by Katherine Crowson.
The scheduler can be easily swapped into existing pipelines like so:

from diffusers import StableDiffusionPipeline, LMSDiscreteScheduler

model_id = "CompVis/stable-diffusion-v1-3-diffusers"
# Use the K-LMS scheduler here instead
scheduler = LMSDiscreteScheduler(beta_start=0.00085, beta_end=0.012, beta_schedule="scaled_linear", num_train_timesteps=1000)
pipe = StableDiffusionPipeline.from_pretrained(model_id, scheduler=scheduler, use_auth_token=True)

Integration test with text-to-image script of Stable-Diffusion

#182 and #186 make sure that DDIM and PNDM/PLMS scheduler yield 1-to-1 the same results as stable diffusion.
Try it out yourself:

In Stable-Diffusion:

python scripts/txt2img.py --prompt "a photograph of an astronaut riding a horse" --n_samples 4 --n_iter 1 --fixed_code --plms

or

python scripts/txt2img.py --prompt "a photograph of an astronaut riding a horse" --n_samples 4 --n_iter 1 --fixed_code

In diffusers:

from diffusers import StableDiffusionPipeline, DDIMScheduler
from time import time
from PIL import Image
from einops import rearrange
import numpy as np
import torch
from torch import autocast
from torchvision.utils import make_grid

torch.manual_seed(42)

prompt = "a photograph of an astronaut riding a horse"
#prompt = "a photograph of the eiffel tower on the moon"
#prompt = "an oil painting of a futuristic forest gives"

# uncomment to use DDIM
# scheduler = DDIMScheduler(beta_start=0.00085, beta_end=0.012, beta_schedule="scaled_linear", clip_sample=False, set_alpha_to_one=False)
# pipe = StableDiffusionPipeline.from_pretrained("CompVis/stable-diffusion-v1-3-diffusers", use_auth_token=True, scheduler=scheduler)  # make sure you're logged in with `huggingface-cli login`

pipe = StableDiffusionPipeline.from_pretrained("CompVis/stable-diffusion-v1-3-diffusers", use_auth_token=True)  # make sure you're logged in with `huggingface-cli login`

all_images = []
num_rows = 1
num_columns = 4
for _ in range(num_rows):
    with autocast("cuda"):
        images = pipe(num_columns * [prompt], guidance_scale=7.5, output_type="np")["sample"]  # image here is in [PIL format](https://pillow.readthedocs.io/en/stable/)
        all_images.append(torch.from_numpy(images))

# additionally, save as grid
grid = torch.stack(all_images, 0)
grid = rearrange(grid, 'n b h w c -> (n b) h w c')
grid = rearrange(grid, 'n h w c -> n c h w')
grid = make_grid(grid, nrow=num_rows)

# to image
grid = 255. * rearrange(grid, 'c h w -> h w c').cpu().numpy()
image = Image.fromarray(grid.astype(np.uint8))

image.save(f"./images/diffusers/{'_'.join(prompt.split())}_{round(time())}.png")

Improvements and bugfixes

Full Changelog: 0.1.3...v0.2.0

Contributors

pcuenca, osanseviero, and 5 other contributors
Assets 2
Loading

0.1.3 Patch release

28 Jul 09:04
@patrickvonplaten patrickvonplaten
0.1.3
b6cadce
Compare
Choose a tag to compare
0.1.3 Patch release

This patch releases refactors the model architecture of VQModel or AutoencoderKL including the weight naming. Therefore the official weights of the CompVis organization have been re-uploaded, see:

Corresponding PR: #137

Please make sure to upgrade diffusers to have those models running correctly: pip install --upgrade diffusers

Bug fixes

  • Fix FileNotFoundError: 'model_card_template.md' #136
Assets 2
Loading

Initial release of 🧨 Diffusers

21 Jul 14:52
@LysandreJik LysandreJik
0.1.2
5311f56
This commit was created on GitHub.com and signed with GitHub’s verified signature. The key has expired.
GPG key ID: 4AEE18F83AFDEB23
Expired
Learn about vigilant mode.
Compare
Choose a tag to compare
Initial release of 🧨 Diffusers

These are the release notes of the 🧨 Diffusers library

Introducing Hugging Face's new library for diffusion models.

Diffusion models proved themselves very effective in artificial synthesis, even beating GANs for images. Because of that, they gained traction in the machine learning community and play an important role for systems like DALL-E 2 or Imagen to generate photorealistic images when prompted on text.

While the most prolific successes of diffusion models have been in the computer vision community, these models have also achieved remarkable results in other domains, such as:

and more.

Goals

The goals of diffusers are:

  • to centralize the research of diffusion models from independent repositories to a clear and maintained project,
  • to reproduce high impact machine learning systems such as DALLE and Imagen in a manner that is accessible for the public, and
  • to create an easy to use API that enables one to train their own models or re-use checkpoints from other repositories for inference.

Release overview

Quickstart:

Diffusers aims to be a modular toolbox for diffusion techniques, with a focus the following categories:

🚄 Inference pipelines

Inference pipelines are a collection of end-to-end diffusion systems that can be used out-of-the-box. The goal is for them to stick as close as possible to their original implementation, and they can include components of other libraries (such as text encoders).

The original release contains the following pipelines:

We are currently working on enabling other pipelines for different modalities. The following pipelines are expected to land in a subsequent release:

  • BDDMPipeline for spectrogram-to-sound vocoding
  • GLIDEPipeline to support OpenAI's GLIDE model
  • Grad-TTS for text to audio generation / conditional audio generation
  • A reinforcement learning pipeline (happening in #105)

⏰ Schedulers

  • Schedulers are the algorithms to use diffusion models in inference as well as for training. They include the noise schedules and define algorithm-specific diffusion steps.
  • Schedulers can be used interchangable between diffusion models in inference to find the preferred tradef-off between speed and generation quality.
  • Schedulers are available in numpy, but can easily be transformed into PyTorch.

The goal is for each scheduler to provide one or more step() functions that should be called iteratively to unroll the diffusion loop during the forward pass. They are framework agnostic, but offer conversion methods which should allow easy conversion to PyTorch utilities.

The initial release contains the following schedulers:

🏭 Models

Models are hosted in the src/diffusers/models folder.

For the initial release, you'll get to see a few building blocks, as well as some resulting models:

  • UNet2DModel can be seen as a version of the recent UNet architectures as shown in recent papers. It can be seen as the unconditional version of the UNet model, in opposition to the conditional version that follows below.
  • UNet2DConditionModel is similar to the UNet2DModel, but is conditional: it uses the cross-attention mechanism in order to have skip connections in its downsample and upsample layers. These cross-attentions can be fed by other models. An example of a pipeline using a conditional UNet model is the latent diffusion pipeline.
  • AutoencoderKL and VQModel are still experimental models that are prone to breaking changes in the near future. However, they can already be used as part of the Latent Diffusion pipelines.

📃 Training example

The first release contains a dataset-agnostic unconditional example and a training notebook:

Credits

This library concretizes previous work by many different authors and would not have been possible without their great research and implementations. We'd like to thank, in particular, the following implementations which have helped us in our development and without which the API could not have been as polished today:

We also want to thank @heejkoo for the very helpful overview of papers, code and resources on diffusion models, available here.

Contributors

hojonathanho, pesser, and 4 other contributors
Assets 2
Loading