ramyma

@yury-dymov I'm facing the same issue; I would like to build scoped data based on the request endpoint and params. As @igorkrup mentioned, creating the meta part is already covered...

> Are you using `experimental_persistent_cond_cache`? I'm not sure how this would be an issue otherwise..? I'm not using `experimental_persistent_cond_cache`, however, if you look at `__init__`: https://github.com/AUTOMATIC1111/stable-diffusion-webui/blob/master/modules/processing.py#L182 ```python self.cached_uc = StableDiffusionProcessing.cached_uc...

> At least I'll suggest > > ```python > p = StableDiffusionProcessingTxt2Img(sd_model=shared.sd_model, **args) > try: > p.scripts = script_runner > p.outpath_grids = opts.outdir_txt2img_grids > p.outpath_samples = opts.outdir_txt2img_samples > > shared.state.begin()...

Can we merge this PR?

I can see that skew style rules are still not working correctly on Android with RN `v0.64.2`. Even the example provided on the [transforms documentation page](https://reactnative.dev/docs/transforms) doesn't work:  However...

I would like to point out that it seems like on Android the skew y operation is doing a bit of rotation as well if you compare it to the...

> Exact same payload works fine with img2img, its only txt2img giving me issues. Looks like you're missing the `image` arg; with img2img it defaults to the main image, but...

Basically the `input_image` property changed to `image` in the controlnet unit; changing the request payload accordingly should fix the issue.

It looks like the controlnet model file you're using is broken, did it download the model again why you try to generate again? Also does it work for you through...

@Haoming02 did you get a chance to look at this thread?