Diffusers documentation

Kandinsky

Diffusers

You are viewing v0.18.2 version. A newer version v0.31.0 is available.

Join the Hugging Face community

and get access to the augmented documentation experience

Collaborate on models, datasets and Spaces

Faster examples with accelerated inference

Switch between documentation themes

to get started

Kandinsky

Overview

Kandinsky inherits best practices from DALL-E 2 and Latent Diffusion, while introducing some new ideas.

It uses CLIP for encoding images and text, and a diffusion image prior (mapping) between latent spaces of CLIP modalities. This approach enhances the visual performance of the model and unveils new horizons in blending images and text-guided image manipulation.

The Kandinsky model is created by Arseniy Shakhmatov, Anton Razzhigaev, Aleksandr Nikolich, Igor Pavlov, Andrey Kuznetsov and Denis Dimitrov. The original codebase can be found here

Usage example

In the following, we will walk you through some examples of how to use the Kandinsky pipelines to create some visually aesthetic artwork.

Text-to-Image Generation

For text-to-image generation, we need to use both KandinskyPriorPipeline and KandinskyPipeline. The first step is to encode text prompts with CLIP and then diffuse the CLIP text embeddings to CLIP image embeddings, as first proposed in DALL-E 2. Let’s throw a fun prompt at Kandinsky to see what it comes up with.

prompt = "A alien cheeseburger creature eating itself, claymation, cinematic, moody lighting"

First, let’s instantiate the prior pipeline and the text-to-image pipeline. Both pipelines are diffusion models.

from diffusers import DiffusionPipeline
import torch

pipe_prior = DiffusionPipeline.from_pretrained("kandinsky-community/kandinsky-2-1-prior", torch_dtype=torch.float16)
pipe_prior.to("cuda")

t2i_pipe = DiffusionPipeline.from_pretrained("kandinsky-community/kandinsky-2-1", torch_dtype=torch.float16)
t2i_pipe.to("cuda")

By default, the text-to-image pipeline use DDIMScheduler, you can change the scheduler to DDPMScheduler

scheduler = DDPMScheduler.from_pretrained("kandinsky-community/kandinsky-2-1", subfolder="ddpm_scheduler")
t2i_pipe = DiffusionPipeline.from_pretrained(
    "kandinsky-community/kandinsky-2-1", scheduler=scheduler, torch_dtype=torch.float16
)
t2i_pipe.to("cuda")

Now we pass the prompt through the prior to generate image embeddings. The prior returns both the image embeddings corresponding to the prompt and negative/unconditional image embeddings corresponding to an empty string.

image_embeds, negative_image_embeds = pipe_prior(prompt, guidance_scale=1.0).to_tuple()

The text-to-image pipeline expects both image_embeds, negative_image_embeds and the original prompt as the text-to-image pipeline uses another text encoder to better guide the second diffusion process of t2i_pipe.

By default, the prior returns unconditioned negative image embeddings corresponding to the negative prompt of "". For better results, you can also pass a negative_prompt to the prior. This will increase the effective batch size of the prior by a factor of 2.

prompt = "A alien cheeseburger creature eating itself, claymation, cinematic, moody lighting"
negative_prompt = "low quality, bad quality"

image_embeds, negative_image_embeds = pipe_prior(prompt, negative_prompt, guidance_scale=1.0).to_tuple()

Next, we can pass the embeddings as well as the prompt to the text-to-image pipeline. Remember that in case you are using a customized negative prompt, that you should pass this one also to the text-to-image pipelines with negative_prompt=negative_prompt:

image = t2i_pipe(
    prompt, image_embeds=image_embeds, negative_image_embeds=negative_image_embeds, height=768, width=768
).images[0]
image.save("cheeseburger_monster.png")

One cheeseburger monster coming up! Enjoy!

The Kandinsky model works extremely well with creative prompts. Here is some of the amazing art that can be created using the exact same process but with different prompts.

prompt = "bird eye view shot of a full body woman with cyan light orange magenta makeup, digital art, long braided hair her face separated by makeup in the style of yin Yang surrealism, symmetrical face, real image, contrasting tone, pastel gradient background"

prompt = "A car exploding into colorful dust"

prompt = "editorial photography of an organic, almost liquid smoke style armchair"

prompt = "birds eye view of a quilted paper style alien planet landscape, vibrant colours, Cinematic lighting"

Text Guided Image-to-Image Generation

The same Kandinsky model weights can be used for text-guided image-to-image translation. In this case, just make sure to load the weights using the KandinskyImg2ImgPipeline pipeline.

Note: You can also directly move the weights of the text-to-image pipelines to the image-to-image pipelines without loading them twice by making use of the components function as explained here.

Let’s download an image.

from PIL import Image
import requests
from io import BytesIO

# download image
url = "https://raw.githubusercontent.com/CompVis/stable-diffusion/main/assets/stable-samples/img2img/sketch-mountains-input.jpg"
response = requests.get(url)
original_image = Image.open(BytesIO(response.content)).convert("RGB")
original_image = original_image.resize((768, 512))

import torch
from diffusers import KandinskyImg2ImgPipeline, KandinskyPriorPipeline

# create prior
pipe_prior = KandinskyPriorPipeline.from_pretrained(
    "kandinsky-community/kandinsky-2-1-prior", torch_dtype=torch.float16
)
pipe_prior.to("cuda")

# create img2img pipeline
pipe = KandinskyImg2ImgPipeline.from_pretrained("kandinsky-community/kandinsky-2-1", torch_dtype=torch.float16)
pipe.to("cuda")

prompt = "A fantasy landscape, Cinematic lighting"
negative_prompt = "low quality, bad quality"

image_embeds, negative_image_embeds = pipe_prior(prompt, negative_prompt).to_tuple()

out = pipe(
    prompt,
    image=original_image,
    image_embeds=image_embeds,
    negative_image_embeds=negative_image_embeds,
    height=768,
    width=768,
    strength=0.3,
)

out.images[0].save("fantasy_land.png")

Text Guided Inpainting Generation

You can use KandinskyInpaintPipeline to edit images. In this example, we will add a hat to the portrait of a cat.

from diffusers import KandinskyInpaintPipeline, KandinskyPriorPipeline
from diffusers.utils import load_image
import torch
import numpy as np

pipe_prior = KandinskyPriorPipeline.from_pretrained(
    "kandinsky-community/kandinsky-2-1-prior", torch_dtype=torch.float16
)
pipe_prior.to("cuda")

prompt = "a hat"
prior_output = pipe_prior(prompt)

pipe = KandinskyInpaintPipeline.from_pretrained("kandinsky-community/kandinsky-2-1-inpaint", torch_dtype=torch.float16)
pipe.to("cuda")

init_image = load_image(
    "https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main" "/kandinsky/cat.png"
)

mask = np.ones((768, 768), dtype=np.float32)
# Let's mask out an area above the cat's head
mask[:250, 250:-250] = 0

out = pipe(
    prompt,
    image=init_image,
    mask_image=mask,
    **prior_output,
    height=768,
    width=768,
    num_inference_steps=150,
)

image = out.images[0]
image.save("cat_with_hat.png")

Interpolate

The KandinskyPriorPipeline also comes with a cool utility function that will allow you to interpolate the latent space of different images and texts super easily. Here is an example of how you can create an Impressionist-style portrait for your pet based on “The Starry Night”.

Note that you can interpolate between texts and images - in the below example, we passed a text prompt “a cat” and two images to the interplate function, along with a weights variable containing the corresponding weights for each condition we interplate.

from diffusers import KandinskyPriorPipeline, KandinskyPipeline
from diffusers.utils import load_image
import PIL

import torch

pipe_prior = KandinskyPriorPipeline.from_pretrained(
    "kandinsky-community/kandinsky-2-1-prior", torch_dtype=torch.float16
)
pipe_prior.to("cuda")

img1 = load_image(
    "https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main" "/kandinsky/cat.png"
)

img2 = load_image(
    "https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main" "/kandinsky/starry_night.jpeg"
)

# add all the conditions we want to interpolate, can be either text or image
images_texts = ["a cat", img1, img2]

# specify the weights for each condition in images_texts
weights = [0.3, 0.3, 0.4]

# We can leave the prompt empty
prompt = ""
prior_out = pipe_prior.interpolate(images_texts, weights)

pipe = KandinskyPipeline.from_pretrained("kandinsky-community/kandinsky-2-1", torch_dtype=torch.float16)
pipe.to("cuda")

image = pipe(prompt, **prior_out, height=768, width=768).images[0]

image.save("starry_cat.png")

Text-to-Image Generation with ControlNet Conditioning

In the following, we give a simple example of how to use KandinskyV22ControlnetPipeline to add control to the text-to-image generation with a depth image.

First, let’s take an image and extract its depth map.

from diffusers.utils import load_image

img = load_image(
    "https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main/kandinskyv22/cat.png"
).resize((768, 768))

We can use the depth-estimation pipeline from transformers to process the image and retrieve its depth map.

import torch
import numpy as np

from transformers import pipeline
from diffusers.utils import load_image


def make_hint(image, depth_estimator):
    image = depth_estimator(image)["depth"]
    image = np.array(image)
    image = image[:, :, None]
    image = np.concatenate([image, image, image], axis=2)
    detected_map = torch.from_numpy(image).float() / 255.0
    hint = detected_map.permute(2, 0, 1)
    return hint


depth_estimator = pipeline("depth-estimation")
hint = make_hint(img, depth_estimator).unsqueeze(0).half().to("cuda")

Now, we load the prior pipeline and the text-to-image controlnet pipeline

from diffusers import KandinskyV22PriorPipeline, KandinskyV22ControlnetPipeline

pipe_prior = KandinskyV22PriorPipeline.from_pretrained(
    "kandinsky-community/kandinsky-2-2-prior", torch_dtype=torch.float16
)
pipe_prior = pipe_prior.to("cuda")

pipe = KandinskyV22ControlnetPipeline.from_pretrained(
    "kandinsky-community/kandinsky-2-2-controlnet-depth", torch_dtype=torch.float16
)
pipe = pipe.to("cuda")

We pass the prompt and negative prompt through the prior to generate image embeddings

prompt = "A robot, 4k photo"

negative_prior_prompt = "lowres, text, error, cropped, worst quality, low quality, jpeg artifacts, ugly, duplicate, morbid, mutilated, out of frame, extra fingers, mutated hands, poorly drawn hands, poorly drawn face, mutation, deformed, blurry, dehydrated, bad anatomy, bad proportions, extra limbs, cloned face, disfigured, gross proportions, malformed limbs, missing arms, missing legs, extra arms, extra legs, fused fingers, too many fingers, long neck, username, watermark, signature"

generator = torch.Generator(device="cuda").manual_seed(43)
image_emb, zero_image_emb = pipe_prior(
    prompt=prompt, negative_prompt=negative_prior_prompt, generator=generator
).to_tuple()

Now we can pass the image embeddings and the depth image we extracted to the controlnet pipeline. With Kandinsky 2.2, only prior pipelines accept prompt input. You do not need to pass the prompt to the controlnet pipeline.

images = pipe(
    image_embeds=image_emb,
    negative_image_embeds=zero_image_emb,
    hint=hint,
    num_inference_steps=50,
    generator=generator,
    height=768,
    width=768,
).images

images[0].save("robot_cat.png")

The output image looks as follow:

Image-to-Image Generation with ControlNet Conditioning

Kandinsky 2.2 also includes a KandinskyV22ControlnetImg2ImgPipeline that will allow you to add control to the image generation process with both the image and its depth map. This pipeline works really well with KandinskyV22PriorEmb2EmbPipeline, which generates image embeddings based on both a text prompt and an image.

For our robot cat example, we will pass the prompt and cat image together to the prior pipeline to generate an image embedding. We will then use that image embedding and the depth map of the cat to further control the image generation process.

We can use the same cat image and its depth map from the last example.

import torch
import numpy as np

from diffusers import KandinskyV22PriorEmb2EmbPipeline, KandinskyV22ControlnetImg2ImgPipeline
from diffusers.utils import load_image
from transformers import pipeline

img = load_image(
    "https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main" "/kandinskyv22/cat.png"
).resize((768, 768))


def make_hint(image, depth_estimator):
    image = depth_estimator(image)["depth"]
    image = np.array(image)
    image = image[:, :, None]
    image = np.concatenate([image, image, image], axis=2)
    detected_map = torch.from_numpy(image).float() / 255.0
    hint = detected_map.permute(2, 0, 1)
    return hint


depth_estimator = pipeline("depth-estimation")
hint = make_hint(img, depth_estimator).unsqueeze(0).half().to("cuda")

pipe_prior = KandinskyV22PriorEmb2EmbPipeline.from_pretrained(
    "kandinsky-community/kandinsky-2-2-prior", torch_dtype=torch.float16
)
pipe_prior = pipe_prior.to("cuda")

pipe = KandinskyV22ControlnetImg2ImgPipeline.from_pretrained(
    "kandinsky-community/kandinsky-2-2-controlnet-depth", torch_dtype=torch.float16
)
pipe = pipe.to("cuda")

prompt = "A robot, 4k photo"
negative_prior_prompt = "lowres, text, error, cropped, worst quality, low quality, jpeg artifacts, ugly, duplicate, morbid, mutilated, out of frame, extra fingers, mutated hands, poorly drawn hands, poorly drawn face, mutation, deformed, blurry, dehydrated, bad anatomy, bad proportions, extra limbs, cloned face, disfigured, gross proportions, malformed limbs, missing arms, missing legs, extra arms, extra legs, fused fingers, too many fingers, long neck, username, watermark, signature"

generator = torch.Generator(device="cuda").manual_seed(43)

# run prior pipeline

img_emb = pipe_prior(prompt=prompt, image=img, strength=0.85, generator=generator)
negative_emb = pipe_prior(prompt=negative_prior_prompt, image=img, strength=1, generator=generator)

# run controlnet img2img pipeline
images = pipe(
    image=img,
    strength=0.5,
    image_embeds=img_emb.image_embeds,
    negative_image_embeds=negative_emb.image_embeds,
    hint=hint,
    num_inference_steps=50,
    generator=generator,
    height=768,
    width=768,
).images

images[0].save("robot_cat.png")

Here is the output. Compared with the output from our text-to-image controlnet example, it kept a lot more cat facial details from the original image and worked into the robot style we asked for.

Kandinsky 2.2

The Kandinsky 2.2 release includes robust new text-to-image models that support text-to-image generation, image-to-image generation, image interpolation, and text-guided image inpainting. The general workflow to perform these tasks using Kandinsky 2.2 is the same as in Kandinsky 2.1. First, you will need to use a prior pipeline to generate image embeddings based on your text prompt, and then use one of the image decoding pipelines to generate the output image. The only difference is that in Kandinsky 2.2, all of the decoding pipelines no longer accept the prompt input, and the image generation process is conditioned with only image_embeds and negative_image_embeds.

Let’s look at an example of how to perform text-to-image generation using Kandinsky 2.2.

First, let’s create the prior pipeline and text-to-image pipeline with Kandinsky 2.2 checkpoints.

from diffusers import DiffusionPipeline
import torch

pipe_prior = DiffusionPipeline.from_pretrained("kandinsky-community/kandinsky-2-2-prior", torch_dtype=torch.float16)
pipe_prior.to("cuda")

t2i_pipe = DiffusionPipeline.from_pretrained("kandinsky-community/kandinsky-2-2-decoder", torch_dtype=torch.float16)
t2i_pipe.to("cuda")

You can then use pipe_prior to generate image embeddings.

prompt = "portrait of a women, blue eyes, cinematic"
negative_prompt = "low quality, bad quality"

image_embeds, negative_image_embeds = pipe_prior(prompt, guidance_scale=1.0).to_tuple()

Now you can pass these embeddings to the text-to-image pipeline. When using Kandinsky 2.2 you don’t need to pass the prompt (but you do with the previous version, Kandinsky 2.1).

image = t2i_pipe(image_embeds=image_embeds, negative_image_embeds=negative_image_embeds, height=768, width=768).images[
    0
]
image.save("portrait.png")

We used the text-to-image pipeline as an example, but the same process applies to all decoding pipelines in Kandinsky 2.2. For more information, please refer to our API section for each pipeline.

Optimization

Running Kandinsky in inference requires running both a first prior pipeline: KandinskyPriorPipeline and a second image decoding pipeline which is one of KandinskyPipeline, KandinskyImg2ImgPipeline, or KandinskyInpaintPipeline.

The bulk of the computation time will always be the second image decoding pipeline, so when looking into optimizing the model, one should look into the second image decoding pipeline.

When running with PyTorch < 2.0, we strongly recommend making use of xformers to speed-up the optimization. This can be done by simply running:

from diffusers import DiffusionPipeline
import torch

t2i_pipe = DiffusionPipeline.from_pretrained("kandinsky-community/kandinsky-2-1", torch_dtype=torch.float16)
t2i_pipe.enable_xformers_memory_efficient_attention()

When running on PyTorch >= 2.0, PyTorch’s SDPA attention will automatically be used. For more information on PyTorch’s SDPA, feel free to have a look at this blog post.

To have explicit control , you can also manually set the pipeline to use PyTorch’s 2.0 efficient attention:

from diffusers.models.attention_processor import AttnAddedKVProcessor2_0

t2i_pipe.unet.set_attn_processor(AttnAddedKVProcessor2_0())

The slowest and most memory intense attention processor is the default AttnAddedKVProcessor processor. We do not recommend using it except for testing purposes or cases where very high determistic behaviour is desired. You can set it with:

from diffusers.models.attention_processor import AttnAddedKVProcessor

t2i_pipe.unet.set_attn_processor(AttnAddedKVProcessor())

With PyTorch >= 2.0, you can also use Kandinsky with torch.compile which depending on your hardware can signficantly speed-up your inference time once the model is compiled. To use Kandinsksy with torch.compile, you can do:

t2i_pipe.unet.to(memory_format=torch.channels_last)
t2i_pipe.unet = torch.compile(t2i_pipe.unet, mode="reduce-overhead", fullgraph=True)

After compilation you should see a very fast inference time. For more information, feel free to have a look at Our PyTorch 2.0 benchmark.

Available Pipelines:

Pipeline	Tasks
pipeline_kandinsky2_2.py	Text-to-Image Generation
pipeline_kandinsky.py	Text-to-Image Generation
pipeline_kandinsky2_2_inpaint.py	Image-Guided Image Generation
pipeline_kandinsky_inpaint.py	Image-Guided Image Generation
pipeline_kandinsky2_2_img2img.py	Image-Guided Image Generation
pipeline_kandinsky_img2img.py	Image-Guided Image Generation
pipeline_kandinsky2_2_controlnet.py	Image-Guided Image Generation
pipeline_kandinsky2_2_controlnet_img2img.py	Image-Guided Image Generation

Diffusers

Kandinsky

Overview

Usage example

Text-to-Image Generation

Text Guided Image-to-Image Generation

Text Guided Inpainting Generation

Interpolate

Text-to-Image Generation with ControlNet Conditioning

Image-to-Image Generation with ControlNet Conditioning

Kandinsky 2.2

Optimization

Available Pipelines:

KandinskyV22Pipeline

class diffusers.KandinskyV22Pipeline

__call__

enable_model_cpu_offload

enable_sequential_cpu_offload

KandinskyV22ControlnetPipeline

class diffusers.KandinskyV22ControlnetPipeline

__call__

enable_model_cpu_offload

enable_sequential_cpu_offload

KandinskyV22ControlnetImg2ImgPipeline

class diffusers.KandinskyV22ControlnetImg2ImgPipeline

__call__

enable_model_cpu_offload

enable_sequential_cpu_offload

KandinskyV22Img2ImgPipeline

class diffusers.KandinskyV22Img2ImgPipeline

__call__

enable_model_cpu_offload

enable_sequential_cpu_offload

KandinskyV22InpaintPipeline

class diffusers.KandinskyV22InpaintPipeline

__call__

enable_model_cpu_offload

enable_sequential_cpu_offload

KandinskyV22PriorPipeline

KandinskyV22PriorEmb2EmbPipeline

class diffusers.KandinskyV22PriorEmb2EmbPipeline

__call__

interpolate

enable_sequential_cpu_offload

KandinskyPriorPipeline

class diffusers.KandinskyPriorPipeline

__call__

interpolate

enable_sequential_cpu_offload

KandinskyPipeline

class diffusers.KandinskyPipeline

__call__

enable_model_cpu_offload

enable_sequential_cpu_offload

KandinskyImg2ImgPipeline

class diffusers.KandinskyImg2ImgPipeline

__call__

enable_model_cpu_offload

enable_sequential_cpu_offload

KandinskyInpaintPipeline

class diffusers.KandinskyInpaintPipeline

__call__

enable_model_cpu_offload

enable_sequential_cpu_offload

call

call

call

call

call

call

call

call

call

call