melmurp

I have a 5070 TI working with this docker, which Cuda version are you on? Support was added in 12.8 I think Tue Jan 27 21:41:33 2026 +-----------------------------------------------------------------------------------------+ | NVIDIA-SMI 580.95.05 Driver Version: 580.95.05 CUDA Version: 13.0 | +-----------------------------------------+------------------------+----------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=========================================================================================| | 0 N/A N/A 2197523 C /comfy/mnt/venv/bin/python3 336MiB | +-----------------------------------------------------------------------------------------+ Tensor compile is busted but that's an issue with cuda itself I believe.

Can happens if you touched the requirements.txt file and it doesn't match git repo anymore... comfyui then refuses to upgrade properly Quick fix is to update comfyui-frontend-package== to higher version then restart comfyui. I think I just deleted the file then did an update and it pulled fresh

I noticed a new TORCH_LOCK field was added to the template and set to required, the README makes it seem like this is optional. If you disable the required flag it will fail to start but if you delete the parameter it comes up fine. Right now this forces the user to set it correctly or delete it from the template which was confusing to me as that's not stated in the README *EDIT: I also notice on fresh install the config.ini network_mode is set to public instead of personal_cloud so you can't install nodes They also moved it so their wiki is outdated.. I found it under ./user/_manager/ not ./user/default/ComfyUI-Manager/

Blackwell cards are only supported by the open source driver on unraid, so 580.

That was it.. updated to 4 and now got latest Appreciate the tip

There was a recent security bulleting pushed out about driver vulnerabilities... I'm unsure what "driver branch" is in use with this plugin and how would you manually update? This is what I see available to me on 7 beta 2 4060 Ti is the card Thanks!

Different AI tools so it varies Example: OpenedAI Speech <process_info> <gpu_instance_id>N/A</gpu_instance_id> <compute_instance_id>N/A</compute_instance_id> <pid>2093547</pid> <type>C</type> <process_name>python</process_name> <used_memory>3100 MiB</used_memory> </process_info> cat /proc/2093547/cmdlin pythonspeech.py

Sorry, some reason I don't get notifications on replies here... Think I got my answer tho... filtered by app name I'm guessing "appssupp":["plex","jellyfin","handbrake","emby","tdarr","unmanic","dizquetv","ersatztv","fileflows","frigate","Threadfin","codeproject","deepstack","nsfminer","shinobipro","foldinghome","compreface"] nvidia-smi is quite verbose so I attached as txt output.txt

I'm not using it to transcode, the memory in use is over 90% and nvidia-smi returns an active application, python in this case, with a pid but that's not being reflected in the dashboard. Here's another shot showing it using more than just the memory.. is Active app only enabled when transcoding is triggered?

Hi, I was curious why the active apps seems to be always empty on the dashboard. If I run nvidia-smi -q I get a process list with the application and vram in use. I have everything enabled in the nvidia section and general Thanks!

I can confirm and thought the same thing... Proton is just Wine but something is happening under the covers with their game engine that seems to really prefer Proton over Wine as I'm seeing better performance as well.

Yeah, I just realized two are vars not ports I have the dumb *edit Copied over world from wine version and it's all running fine... I'm seeing ~10-15% less cpu with the proton version but yet to have more than a few players (2) to test with.

This happens if I remove and readd them

When I started the docker it tossed an error saying port already allocated and the docker cmd had 4 ports being mapped.. two dupes *edit I had to delete all four and readd them since I use different ports so perhaps that causes it?

I have it running on a dual E5-2630L v2 @ 2.40GHz and it barely touches 35% usage but nose dives after 4 ppl join... yet CPU stays low. I was told... get a newer CPU... which seemed odd as.. well.. it's not bottlenecking on cpu. So now I'm trying it on a Ryzen 5 3600 to see if "newer cpu" does anything. Grabbing the proton one now... something I did see was you have multiples UDP port entries.. guessing one was for config and other for actual ports? Docker tried to allocate them twice so I just removed one set.