Integrate Stream Diffusion

Use Stream diffusion to generate image or as a modifer inside Smode

StreamDiffusion is currently not available on RTX 5XXX GPUs because PyTorch does not yet support them. You can track the progress of this issue on the official PyTorch GitHub repository: https://github.com/pytorch/pytorch/issues/145949

Instalation

First Enable Windows Long Path

The first time you use the Stream Diffusion Modifier(beta) or Stream Diffusion Generator (beta) on a computer you need to install the dependencies by click the installation trigger.

the instalation take can take lot of time ( 20 min )

Optionaly you can install TensorRT, tensorRT will allow you to have faster generation in trade of control ( changing some parameter will have no effect until a reload )

TensorRT

The tensorRT aceleration require you to have installed tensorRT
When you load a StreamDiffusion with new parameters you need to wait for the model to be compiled, this can take multiple minute

Debug

Inside the log folder of Smode there is a log for the Streamdiffusion instance, if the installation fail or the generation don’t work you need to look for error inside the log

Unistallation

you can use the trigger button that is inside the Engine preference pannel under StreamDiffusion configuration to uninstall the plugin.
or go inside the C:\ProgramData\SmodeTech\smode Edition\Packages folder