ControlNet-v1-1-nightly VS diffusers

1- https://github.com/lllyasviel/ControlNet-v1-1-nightly/issues/89

Install ControlNet and download the models you want to use (canny/depth/openpose should be enough for this): https://github.com/lllyasviel/ControlNet-v1-1-nightly

you can find the previous version on their git and if it's a previous version of v1.1 then you probably have to search for the right branch on the new git version and download that

Stable Diffusion: https://huggingface.co/runwayml/stable-diffusion-v1-5 with ControlNet extension: https://github.com/lllyasviel/ControlNet-v1-1-nightly running on Automatic1111 web UI: https://github.com/AUTOMATIC1111/stable-diffusion-webui

Sure thing, after testing midjourney a bit I found out that yhe quality of images produced is best but you have zero control on over what is produced. The big breakthrough here is ControlNet which is a Stable Diffusion extension that makes you control the initial noise based on image inputs (or at least this is what i understand) more on it here: https://github.com/lllyasviel/ControlNet-v1-1-nightly

https://github.com/lllyasviel/ControlNet-v1-1-nightly/issues/61 it seems get removed as duplicated.

I'm missing something about the random factor, from the sample code from https://github.com/huggingface/diffusers/blob/main/README.md

Hey people, I noticed that combining these two methods in a single forward pass increases the controllability of the generation quite a bit. I was kind of puzzled that sometimes ControlNet yielded better results than T2IAdapter for some cases, and sometimes it was the other way around, so I decided to test both at the same time, and results were quite nice. Some visuals and more motivation here: https://github.com/huggingface/diffusers/issues/5847 And it was already merged here: https://github.com/huggingface/diffusers/pull/5869

Open Parti Prompts: The better way to evaluate diffusion models (repo)

You have disabled the safety checker for by passing `safety_checker=None`. Ensure that you abide to the conditions of the Stable Diffusion license and do not expose unfiltered results in services or applications open to the public. Both the diffusers team and Hugging Face strongly recommend to keep the safety filter enabled in all public facing circumstances, disabling it only for use-cases that involve analyzing network behavior or auditing its results. For more information, please have a look at https://github.com/huggingface/diffusers/pull/254 .

For developers, TensorRT support also exists for the diffusers library via community pipelines. [1] It's limited, but if you're only supporting a subset of features, it can help.

In general, these insane speed boosts comes at the cost of bleeding edge features.

[1] https://github.com/huggingface/diffusers/blob/28e8d1f6ec82a6...

I was training sdxl UNET base model, with the diffusers library, which was going great until around step 210k when the weights suddenly turned back to their original values and stayed that way. I also tried with the ema version, which didn't change at all. I also looked at the tensor's weight values directly which confirmed my suspicions.

Merging LoRAs is essentially taking a weighted average of the LoRA adapter weights. It's more common in other UIs.

diffusers is working on a PR for it: https://github.com/huggingface/diffusers/pull/4473