StreamDiffusionTD 0.1.11 🚀🎮🖼️ [ TOX ]
Changelog / announcement here >>> https://www.patreon.com/posts/104339324/
Upgrade Guide
Similar to previous version, just copy and paste your Basefolder parameter ( and depending on your setup change other settings in the operator) and it should load your previous working model list. Once you select a Main SD Model, you should be ready to go.
Comments (20)
Hi!!! Any updates regarding the new TD versions ? Mine is crashing often after 11700s
Works with python 12?
Hi. Is there anyway to put negative prompt?
It says that CUDA can't be found even though i downlaoded the correct version!
missing cuda-python in the env. I was able to fix by diving into the main comp, then the streamdiffusionTD comp, then look for the 'requirements' text dat. add 'cuda-python'. then pulse the 2. Install (venv + all req)
i get “assertion error: SRE module mismatch” when trying to install venv (all req) any help on this?
had previously gotten it to work on m2
hello, May I kindly ask for help, I am getting this error: ControlNet model: select_folder_to_create_dropdown Repository Not Found for url: https://huggingface.co/select_folder_to_create_dropdown/resolve/main/. Please make sure you specified the correct `repo_id` and `repo_type`. If you are trying to access a private or gated repo, make sure you are authenticated. Invalid username or password.
anyway, i installed those manually,now its working
which did you install manually? i have3.10.9 but it doesn't work
Trying to do a fresh install on another machine, and I noticed the "view installation guide" is no longer working (it opens up a black window)
Also getting blank screen. I grabbed the nodes for this feature from 0.1.9.1 that were working and pasted them into the new file and they still aren't working.
my python version is Python 3.10.11
i'm just getting this error ERROR: Could not find a version that satisfies the requirement torch==2.1.0 (from versions: none) ERROR: No matching distribution found for torch==2.1.0
What hardware should I be focusing on for the best performance? I'm assuming this is all local GPU powered? If so, should we be using gaming cards, quadro cards, or ai cards? Does having a CPU with AI cores matter? Does system memory and CPU speed come into play? With a 4090 12fps is a bit low and I'm trying to find what hardware this utilizes the most so I can build a server dedicated just to this one task. Thanks for any insight you can send my way and well done! <3
hello i have some issue when i press the start stream, the cmd was pop out with this Text " File "C:\user\download\Stream Diff\StreamDiffussion\streamdiffusionTD\main_sdtd.py", Line 17, in from torchvision.transform.functional import to_tensor ModuleNotFoundError: No module named 'torchvision' can you help me? cause i cant start stream yet
This can occur if you use the latest Python version. You will need to install Python 3.10.9 - https://www.python.org/downloads/release/python-3109/
Thanks Ethan, My thought is StreanFuion needs launch conditions to select one of the GPUs, but I dont think it has that yet. Then I need the StreamDiffusionTD set manualy Sucribe to one of those streams. - As of now I can 2x StreamDiffusion instances, but TD trys to grab both frames!
@DotSim Want to Thank you soo much for the Epic tools! I have a question? I've got two 4090s and I really like to composite two StreamDiffusion streams. I would love to get a perimeter to set GPU0 or GPU1. I absolutely love the moving backgrounds it does, and I really like the characters I can get if you tune the settings. Unfortunately, each of these optimizations is different settings. So composing the two videos would make the most sense. - Is this possible?
You might be able to launch TD with GPU Affinity arguments to get it to prefer a certain GPU! I am not sure if SD plugin chooses its own internally via python or not. https://docs.derivative.ca/Using_Multiple_Graphic_Cards