hlky

* [UNet2DConditionModel](https://github.com/facebookincubator/AITemplate/blob/main/examples/05_stable_diffusion/src/modeling/unet_2d_condition.py) and [UNet blocks](https://github.com/facebookincubator/AITemplate/blob/main/examples/05_stable_diffusion/src/modeling/unet_blocks.py) need updating for architecture changes and for extra conditionings * [CLIP](https://github.com/facebookincubator/AITemplate/blob/main/examples/05_stable_diffusion/src/modeling/clip.py) needs to output correct hidden state layer, also for bigG, text projection and pooled...

@JJHu1993 I have sdxl support in my private projects. When I have time I will bring those changes here.

@CyberTimon I'll do it today. * CLIP changes are done, but need testing for numerical accuracy, also, there is a workaround for the limitations of `dynamic_slice` with regards to pooled...

[compile_controlnet.py](https://github.com/facebookincubator/AITemplate/blob/main/examples/05_stable_diffusion/scripts/compile_controlnet.py#L70) is using `lllyasviel/sd-controlnet-canny`. `hidden_dim` is not passed to [`compile_controlnet`](https://github.com/facebookincubator/AITemplate/blob/bd5ba64d9d2244356a5fc3c75238d43d7bffc155/examples/05_stable_diffusion/scripts/compile_controlnet.py#L73-L82). This results in dimension mismatch for [text_embeddings](https://github.com/facebookincubator/AITemplate/blob/bd5ba64d9d2244356a5fc3c75238d43d7bffc155/examples/05_stable_diffusion/src/compile_lib/compile_controlnet.py#L99-L101) as the config is edited for v2. * Change [`lllyasviel/sd-controlnet-canny` to `thepowefuldeez/sd21-controlnet-canny`](https://github.com/facebookincubator/AITemplate/blob/bd5ba64d9d2244356a5fc3c75238d43d7bffc155/examples/05_stable_diffusion/scripts/compile_controlnet.py#L70) (or...

Ran into a similar issue while working on RRDBNet. ``` def forward(self, x): x1 = self.lrelu(self.conv1(x)) x2 = self.lrelu(self.conv2(self.cat((x, x1), 3))) x3 = self.lrelu(self.conv3(self.cat((x, x1, x2), 3))) x4 = self.lrelu(self.conv4(self.cat((x,...

[Missing sampler from `StableDiffusionInput`](https://github.com/Stability-AI/StableStudio/blob/5bc209cbce2139c8560b5378a395fecc42991513/packages/stablestudio-plugin/src/Plugin.ts#L213C37-L234) + no UI element for sampler

Let me know if there are any changes or anything to add for this initial version

> `getStableDiffusionDefaultCount` appear to be ignored by the ui Can confirm, found the same with ComfyUI plugin, [fix](https://github.com/Stability-AI/StableStudio/pull/36/files#diff-759ef80471a749ceee594b416a9d5958225baf0cb6484946e5381f5882e16d6f) is included in that PR

I have fixed compilation of Stable Diffusion demo on my local branch `X:\meta\AITemplate\examples\05_stable_diffusion>python scripts\demo.py` ``` INFO:aitemplate.backend.build_cache_base:Build cache disabled [06:06:22] X:\meta\AITemplate\examples\05_stable_diffusion\tmp\CLIPTextModel\model_container.cu:67: Device Runtime Version: 12010; Driver Version: 12010 [06:06:22] X:\meta\AITemplate\examples\05_stable_diffusion\tmp\CLIPTextModel\model_container.cu:81: Hardware...

@cruhl here you go, there's some code for "Extras" in the UI that I hadn't pushed yet too Note that ComfyUI needs to be launched with `--enable-cors-header`