yayitazale

Check the changelog breaking change part and if there is anything in your config that must be changed, read the docs to understand how to change it. Then update and check the logs. If there is no problem, then you can also do the same procedure but for the new features that you can add.

It is not mandatory, but I'll recommend it as the template has new entries now (especially for the ones that use a Nvidia card as a detector because the tensor-rt models are now generated in the frigate container itself).

Recheck the docs and the changelog and if you still have issues, ask on github, the template is working good and it is working good for me.

You should open a issue on github then...

Ok can you open a issue in github then? https://github.com/blakeblackshear/frigate/issues And you can quote me @yayitazale Thanks

Can you share how are you launching the container and the config? Are you using the card in another container or pass-through to a VM?

Did you have a db path definition in the config file?

This is pointed out in the changelog:

v13.0 has been released today to stable channel as it has been tested in many beta releases and release candidates and it is indeed stable, so you should reads the changelog and the updated docs to find out what changes you need to make to your settings. Also it is recommended to ensure that you are using all the required variables in the template or reinstall it from scratch.

Not sure. ask here: https://github.com/blakeblackshear/frigate/discussions?discussions_q=is%3Aopen

can't help you with such little info.

PD: the stream feed of the frontend is indeed the one of the detect role. If you have the chance to use a hwaccel card I suggest you to use the restreaming ability to create a internal feed of the main high res stream and use it for both record and detect, you shouldn't see a big difference of CPU usage.

Did you set up the corresponding input path and roles on each camera? https://docs.frigate.video/configuration/cameras#setting-up-camera-inputs If you want to use two different stream inputs one for detection and one for recording you should have something like the example of the link, with one role per stream path. PD: the stream feed of the frontend is indeed the one of the detect role. If you have the chance to use a hwaccel card I suggest you to use the restreaming ability to create a internal feed of the main high res stream and use it for both record and detect, you shouldn't see a big difference of CPU usage.

No idea, please ask on github as it looks a develop error, I can't help you, you are doing it in the right way.

Maybe you have to download the sh again or try another path. Anyway, don't waste too much time on this, the next frigate versión witch will be released very soon will have the tensor model integrated inside the container, so what you are doing will be in vain. I suggest you to wait some weeks and use this feature in the frigate v13.

Do you have the /mnt/user/appdata/trt-models folder mapped already to frigate and frigate running? If yes, please stop frigate, run this container and then restar frigate.

No. There are 2 folders that need to be bind to the unraid host and these are the corresponding suggested binding points that are presented as default in the template: Frigate -> Unraid /config -> /mnt/user/appdata/frigate /media/frigate -> /mnt/user/Media/frigate If you want to use another binding folders you should create it/them before using it.

You should use HACS to install it. https://github.com/blakeblackshear/frigate-hass-integration

Ok, you should ask on github this is out of my knowledge

What are the go2rtc logs?

First of all, look at the logs both in Frigate and WhosAtMyFeeder. Recheck the configuration of WhosAtMyFeeder and the modifications needed on Frigate on each camera setup to work alongside WhosAtMyFeeder. Cheack if the snapshots of frigate are being created properly. Check if bird detections are beeing recieved by WhosAtMyFeeder. If all those things are working, check if the camera resolutions/snapshot resolutions and camera stream formats are yet correct after that upgrade of the camera firmware.

With such little info I can't help you...

preset-record-ubiquity -> preset-record-ubiquiti

https://docs.frigate.video/configuration/camera_specific#unifi-protect-cameras

Did you change anything? Is there anything in the logs of WhosAtMyFeeder?