*Feature update!

high res rendering is now very coherent and high quality with very little overhead. Just set your gen to higher than 512x512 and you will get a very nice render at selected res. You will also get the initial render to use as img2img target if you want later.

Dual Prompting and Split Prompting.

So this is interesting. This only works currently with high res renders (anything above 512x512) but if you want to get more coherence and diversity from your renders you can now prompt in a new way.

Dual Prompt: This is the most basic prompt style. Two prompts seperated by a :

Example: Cartoon still of emma watson as belle in beauty and the beast:movie still of emma watson in beauty and the beast

this is run the first prompt first then the second pass will use the second prompt. They can be anything you want, but it's best to at least have commonality, but it's not necessary.

Split Prompt: Split prompting allows for large diversity. It uses a wildcard to delinate your prompts.

Example: movie still of * belle in beauty and the beast:emma watson as

This will run the first prompt as: movie still of belle in beauty and the beast

and the second prompt will be: movie still of emma watson as belle in beauty and the beast

by doing this it won't attempt to wedge the person in the shot and will result in my more diverse outputs with more interesting poses and actions.

Split Dual Prompt: the final new method is a way of saving some typing time. A split dual prompt will take a wild card and replace it in both prompts

Example:

• of emma watson in beauty and the beast:cartoon still:movie still

this will produce the exact same prompts as the first dual prompt, but it takes less typing. You can use this wildcard anywhere and it will replace the first phrase in the first prompt and the second in the second. Get creative with it!

---

This is a bit of a diversion from the standard stable diffusion build. It makes txt2img and img2img into a module so it can load the model into memory at launch. You need to download the weights for this to work. Go to https://huggingface.co/CompVis and set up an account to download the weights. You have to agree to terms. Put the weight checkpoint file into ./stable-diffusion-main/models/ldm/stable-diffusion-v1 and it will pick it up. You can have as many checkpoints as you like and it will reload the model if you change to a new one. Have fun!

Some new additions are, Animation, just render an image then press the animate button. If you don't set a project or frame length you will get 60 frames to an untitled project in the animate output. Added the ability to use textual inversion .pt files. Add a folder called embedding in the root dir and drop the embeddings.pt file and name it your keyword. Thanks to Automatic for that.

modules used in GUI: PyQt5 itertools subprocess easygui os random math PIL qdarkstyle

Please go to https://ommer-lab.com/files/latent-diffusion/sr_bsr.zip and extract the ckpt file and place it in models/ldm/bsr_sr/ to use Latent_sr for BSRGaN go into the BSRGAN_main/model_zoo folder and check readme for links to access checkpoint files

https://www.youtube.com/watch?v=7VSi8NufW64

Stable Diffusion

Stable Diffusion was made possible thanks to a collaboration with Stability AI and Runway and builds upon our previous work:

High-Resolution Image Synthesis with Latent Diffusion Models
Robin Rombach*, Andreas Blattmann*, Dominik Lorenz, Patrick Esser, Björn Ommer

which is available on GitHub.

Stable Diffusion is a latent text-to-image diffusion model. Thanks to a generous compute donation from Stability AI and support from LAION, we were able to train a Latent Diffusion Model on 512x512 images from a subset of the LAION-5B database. Similar to Google's Imagen, 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 this section below and the model card.

Requirements

A suitable conda environment named ldm can be created and activated with:

conda env create -f environment.yaml
conda activate ldm

You can also update an existing latent diffusion environment by running

conda install pytorch torchvision -c pytorch
pip install transformers==4.19.2
pip install -e .

Stable Diffusion v1

Stable Diffusion v1 refers to a specific configuration of the model architecture that uses a downsampling-factor 8 autoencoder with an 860M UNet and CLIP ViT-L/14 text encoder for the diffusion model. The model was pretrained on 256x256 images and then finetuned on 512x512 images.

Note: Stable Diffusion v1 is a general text-to-image diffusion model and therefore mirrors biases and (mis-)conceptions that are present in its training data. Details on the training procedure and data, as well as the intended use of the model can be found in the corresponding model card. Research into the safe deployment of general text-to-image models is an ongoing effort. To prevent misuse and harm, we currently provide access to the checkpoints only for academic research purposes upon request. This is an experiment in safe and community-driven publication of a capable and general text-to-image model. We are working on a public release with a more permissive license that also incorporates ethical considerations.

Request access to Stable Diffusion v1 checkpoints for academic research

Weights

We currently provide three checkpoints, sd-v1-1.ckpt, sd-v1-2.ckpt and sd-v1-3.ckpt, which were trained as follows,

• sd-v1-1.ckpt: 237k steps at resolution 256x256 on laion2B-en. 194k steps at resolution 512x512 on laion-high-resolution (170M examples from LAION-5B with resolution >= 1024x1024).
• sd-v1-2.ckpt: Resumed from sd-v1-1.ckpt. 515k steps at resolution 512x512 on "laion-improved-aesthetics" (a subset of laion2B-en, filtered to images with an original size >= 512x512, estimated aesthetics score > 5.0, and an estimated watermark probability < 0.5. The watermark estimate is from the LAION-5B metadata, the aesthetics score is estimated using an improved aesthetics estimator).
• sd-v1-3.ckpt: Resumed from sd-v1-2.ckpt. 195k steps at resolution 512x512 on "laion-improved-aesthetics" and 10% dropping of the text-conditioning to improve classifier-free guidance sampling.

Evaluations with different classifier-free guidance scales (1.5, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0) and 50 PLMS sampling steps show the relative improvements of the checkpoints:

Text-to-Image with Stable Diffusion

Stable Diffusion is a latent diffusion model conditioned on the (non-pooled) text embeddings of a CLIP ViT-L/14 text encoder.

After obtaining the weights, link them

mkdir -p models/ldm/stable-diffusion-v1/
ln -s <path/to/model.ckpt> models/ldm/stable-diffusion-v1/model.ckpt

and sample with

python scripts/txt2img.py --prompt "a photograph of an astronaut riding a horse" --plms

By default, this uses a guidance scale of --scale 7.5, Katherine Crowson's implementation of the PLMS sampler, and renders images of size 512x512 (which it was trained on) in 50 steps. All supported arguments are listed below (type python scripts/txt2img.py --help).

usage: txt2img.py [-h] [--prompt [PROMPT]] [--outdir [OUTDIR]] [--skip_grid] [--skip_save] [--ddim_steps DDIM_STEPS] [--plms] [--laion400m] [--fixed_code] [--ddim_eta DDIM_ETA] [--n_iter N_ITER] [--H H] [--W W] [--C C] [--f F] [--n_samples N_SAMPLES] [--n_rows N_ROWS]
                  [--scale SCALE] [--from-file FROM_FILE] [--config CONFIG] [--ckpt CKPT] [--seed SEED] [--precision {full,autocast}]

optional arguments:
  -h, --help            show this help message and exit
  --prompt [PROMPT]     the prompt to render
  --outdir [OUTDIR]     dir to write results to
  --skip_grid           do not save a grid, only individual samples. Helpful when evaluating lots of samples
  --skip_save           do not save individual samples. For speed measurements.
  --ddim_steps DDIM_STEPS
                        number of ddim sampling steps
  --plms                use plms sampling
  --laion400m           uses the LAION400M model
  --fixed_code          if enabled, uses the same starting code across samples
  --ddim_eta DDIM_ETA   ddim eta (eta=0.0 corresponds to deterministic sampling
  --n_iter N_ITER       sample this often
  --H H                 image height, in pixel space
  --W W                 image width, in pixel space
  --C C                 latent channels
  --f F                 downsampling factor
  --n_samples N_SAMPLES
                        how many samples to produce for each given prompt. A.k.a. batch size
  --n_rows N_ROWS       rows in the grid (default: n_samples)
  --scale SCALE         unconditional guidance scale: eps = eps(x, empty) + scale * (eps(x, cond) - eps(x, empty))
  --from-file FROM_FILE
                        if specified, load prompts from this file
  --config CONFIG       path to config which constructs model
  --ckpt CKPT           path to checkpoint of model
  --seed SEED           the seed (for reproducible sampling)
  --precision {full,autocast}
                        evaluate at this precision

Note: The inference config for all v1 versions is designed to be used with EMA-only checkpoints. For this reason use_ema=False is set in the configuration, otherwise the code will try to switch from non-EMA to EMA weights. If you want to examine the effect of EMA vs no EMA, we provide "full" checkpoints which contain both types of weights. For these, use_ema=False will load and use the non-EMA weights.

Image Modification with Stable Diffusion

By using a diffusion-denoising mechanism as first proposed by SDEdit, the model can be used for different tasks such as text-guided image-to-image translation and upscaling. Similar to the txt2img sampling script, we provide a script to perform image modification with Stable Diffusion.

The following describes an example where a rough sketch made in Pinta is converted into a detailed artwork.

python scripts/img2img.py --prompt "A fantasy landscape, trending on artstation" --init-img <path-to-img.jpg> --strength 0.8

Here, strength is a value between 0.0 and 1.0, that controls the amount of noise that is added to the input image. Values that approach 1.0 allow for lots of variations but will also produce images that are not semantically consistent with the input. See the following example.

Input

Outputs

This procedure can, for example, also be used to upscale samples from the base model.

Comments

• Our codebase for the diffusion models builds heavily on OpenAI's ADM codebase and https://github.com/lucidrains/denoising-diffusion-pytorch. Thanks for open-sourcing!

• The implementation of the transformer encoder is from x-transformers by lucidrains.

BibTeX

@misc{rombach2021highresolution,
      title={High-Resolution Image Synthesis with Latent Diffusion Models},
      author={Robin Rombach and Andreas Blattmann and Dominik Lorenz and Patrick Esser and Björn Ommer},
      year={2021},
      eprint={2112.10752},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}