ifsheldon

+1. Also waiting for an official implementation.

@leezenn Thanks for the suggestion. How did you install streamdiffusion library? I guess in installation Step 3, we need to remove `[tensorrt]` right? Do we need to do extra steps?

@leezenn Thanks a lot! I've successfully run it. But I wonder if you can run it with `sfast`? I don't know what it is. I cannot find it anywhere, in...

+1. Now `diffusers` 0.25 supports ControlNet-XS. Probably that implementation can be a reference.

I believe it's a Lora of controlnet-depth, except their effort on edit consistency. We need a way to apply Lora to controlnet. Do you think it's possible? @comfyanonymous

FYI, we've published the full weights [here](https://github.com/GempollAI/LooseControlNet). You can use it like normal ControlNet.

`wgpu 0.19` is just released. We just need a few minor changes to remove the last bit of unsafe code in the tutorial when creating a `Surface`. It works perfectly...

@tgross35 Thanks for the heads-up. How about async methods? I have the following code in my `GfxStates::new()` ```rust let instance = wgpu::Instance::default(); let surface = instance.create_surface(window).expect("Failed to create surface"); //...

Just FYI, I updated my code to wgpu 0.20, which was released 5 hrs ago and is very easy to migrate to. Just follow rustc's error messages.

Using `wgpu::Surface` with `Window` of `winit>0.30.0` is a bit tricky, which may cause self-reference. You can refer to https://github.com/gfx-rs/wgpu/discussions/6005#discussioncomment-10105507 and [my commit](https://github.com/nanovis/Wenderer/commit/ec0f17a176ddd07ded7569f668d6793a0b3c60a3) that updates my renderer to get an idea...