guarddog13

> My Nano does pretty quick. Currently I have Frigate in a RPi4 8gig with a 64 bit version of Raspian, docker, and docker-compose. I have a Coral TPU assigned...

I did it as a docker run command. I have both my frigate and compreface running on my gpu. The cpu version crashes my RAM periodically and the gpu version...

Frigate GPU with Compreface GPU is near instant as we walk up to the door. I dont think I need a TPU using my GPU. Running it this way takes...

Now I wonder what I'm doing wrong for Deepstack:GPU to fail. I like Compreface but miss Deepstack and CF together.

I always struggle with docker-compose unless I have a template lol. It took forever for me to try to get Frigate setup with compose and I ended up making it...

Thank you so much! It would be greatly appreciated especially if I can get it running on the GPU with Frigate and Compreface. Even with those both on that my...

I see you respond via email and not sure if you see edits so please check as I added a paragraph with a question for you. Thank you!

I thought I figured it out and glad I have more RAM coming lol. Increased my swap space and it quit crashing. I also now have a working docker-compose. Now...

I have done both altering my detect settings in DT and rebuilding/restarting DT. I can get DT to train fine with the CPU version but once I switch to GPU...

I'm running it directly on my own Linux Laptop. It doesn't matter where I try to post for deepstack I'm getting the same error. The errors say that its timing...