[Detector Support]: No events or dections with Nvidia

[KATPWR]

Describe the problem you are having

My Frigate installation sees motion, but never makes any events or attempts to do object detection. In the Debug page I can see it detects motion putting boxs around any motion currently.

Version

0.13.2-6476F8A

Frigate config file

mqtt:
  enabled: false

# Include all cameras by default in Birdseye view
birdseye:
  enabled: true
  mode: continuous

detectors:
  nvidia_2070:
    type: tensorrt

objects:
  track:
    - person
    - car

record:
  enabled: true
  retain:
    days: 14
    mode: all
  events:
    retain:
      default: 30
      mode: motion

model:
  path: /config/model_cache/tensorrt/yolov7-640.trt
  input_tensor: nchw
  input_pixel_format: rgb
  width: 640
  height: 640

cameras:
  front_door_cam:
    ffmpeg:
      hwaccel_args: preset-nvidia-h264
      inputs:
        #HIGH RES
      - path: rtsp://admin:K24Turb0@192.168.0.199:554//h264Preview_01_main
        #- "ffmpeg:http://reolink_ip/flv?port=1935&app=bcs&stream=channel0_main.bcs&user=username&password=redacted#video=copy#audio=copy#audio=opus"
          #input_args: preset-rtsp-restream
        roles:
        - record

      - path: rtsp://admin:K24Turb0@192.168.0.199:554//h264Preview_01_sub
        #- "ffmpeg:http://reolink_ip/flv?port=1935&app=bcs&stream=channel0_ext.bcs&user=username&password=redacted"
          #input_args: preset-rtsp-restream
        roles:
        - detect
        - rtmp
    detect:
      enabled: true
      #width: 640
      #height: 480
      #fps: 5
      stationary:
        #interval: 50
        threshold: 50
    objects:
      track:
      - person
      - car

    zones:
      zone_0:  #drive way
        coordinates: 406,292,640,337,640,480,574,480,479,427,474,401,447,342,372,297
      zone_1:  #walkway
        coordinates: 471,423,570,480,199,480,290,391,450,339,454,357
      zone_2:  #front yard
        coordinates: 224,438,295,394,444,344,361,296,172,304,0,287
      zone_3:  #side yard
        coordinates: 442,294,530,318,578,298,527,287
      zone_4:  #street
        coordinates: 410,272,420,267,570,253,600,259,540,269,640,286,640,299,509,287,193,303,0,286,0,269
    motion:
      mask:
      - 0,271,134,277,308,276,399,272,335,252,370,253,423,259,530,263,575,256,605,252,566,271,640,287,640,0,0,0
motion:
  mask:
  - 0,273,134,277,308,276,399,272,335,252,370,253,423,259,530,263,575,256,605,252,566,271,640,287,640,0,0,0

snapshots: # <----- Enable snapshots
  enabled: true

docker-compose file or Docker CLI command

version: "3.9"
services:
  frigate:
    container_name: frigate
    privileged: true # this may not be necessary for all setups
    restart: unless-stopped
    image: ghcr.io/blakeblackshear/frigate:stable-tensorrt
    shm_size: "64mb" # update for your cameras based on calculation above
    deploy: # NVIDIA GPU
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: 1
              capabilities: [gpu]
    #devices:
      #- /dev/bus/usb:/dev/bus/usb # passes the USB Coral, needs to be modified for other versions
      #- /dev/apex_0:/dev/apex_0 # passes a PCIe Coral, follow driver instructions here https://coral.ai/docs/m2/get-started/#2a-on-linux
      #- /dev/dri/renderD128 # for intel hwaccel, needs to be updated for your hardware
    volumes:
      - /etc/localtime:/etc/localtime:ro #TZ
      - /data/home/frigate:/media/frigate
      - /home/local/compose/frigate/config/config.yml:/config/config.yml
      - /home/local/compose/frigate/config:/config

      - type: tmpfs # Optional: 1GB of memory, reduces SSD/SD Card wear
        target: /tmp/cache #change to ram useage
        tmpfs:
          size: 1000000000
    ports:
      - "5000:5000"
      - "8554:8554" # RTSP feeds
      - "8555:8555/tcp" # WebRTC over tcp
      - "8555:8555/udp" # WebRTC over udp
    networks:
      - local
    environment:
      #- PUID: $FRIGATE
      #- GUID: $FRIGATE
      - FRIGATE_RTSP_PASSWORD="REDACTED"
      - YOLO_MODELS=yolov4-608,yolov7-640,yolov7-320 #generates models for detectors
      - CUDA_MODULE_LOADING=LAZY
      #- USE_FP16=True
    labels:
        #TRAEFIK: REDACTED

networks:
  local:
    external: true

Relevant log output

2024-05-24 12:57:23.896486205  [INFO] Preparing Frigate...
2024-05-24 12:57:23.902814238  [INFO] Starting Frigate...
2024-05-24 12:57:24.756658043  [2024-05-24 12:57:24] frigate.app                    INFO    : Starting Frigate (0.13.2-6476f8a)
2024-05-24 12:57:30.092140941  [2024-05-24 12:57:30] peewee_migrate.logs            INFO    : Starting migrations
2024-05-24 12:57:30.094298116  [2024-05-24 12:57:30] peewee_migrate.logs            INFO    : There is nothing to migrate
2024-05-24 12:57:30.098557200  [2024-05-24 12:57:30] frigate.app                    INFO    : Recording process started: 610
2024-05-24 12:57:30.099917412  [2024-05-24 12:57:30] frigate.app                    INFO    : go2rtc process pid: 102
2024-05-24 12:57:30.112924182  [2024-05-24 12:57:30] detector.nvidia_2070           INFO    : Starting detection process: 619
2024-05-24 12:57:30.228105550  [2024-05-24 12:57:30] frigate.app                    INFO    : Output process started: 621
2024-05-24 12:57:30.228169281  [2024-05-24 12:57:30] frigate.app                    INFO    : Camera processor started for front_door_cam: 628
2024-05-24 12:57:30.228200290  [2024-05-24 12:57:30] frigate.app                    INFO    : Capture process started for front_door_cam: 630
2024-05-24 12:57:30.246022539  [2024-05-24 12:57:30] frigate.detectors.plugins.tensorrt INFO    : Loaded engine size: 71 MiB
2024-05-24 12:57:30.343224559  [2024-05-24 12:57:30] frigate.detectors.plugins.tensorrt INFO    : [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +6, GPU +8, now: CPU 198, GPU 357 (MiB)
2024-05-24 12:57:30.346623039  [2024-05-24 12:57:30] frigate.detectors.plugins.tensorrt INFO    : [MemUsageChange] Init cuDNN: CPU +1, GPU +10, now: CPU 199, GPU 367 (MiB)
2024-05-24 12:57:30.353476118  [2024-05-24 12:57:30] frigate.detectors.plugins.tensorrt INFO    : [MemUsageChange] TensorRT-managed allocation in engine deserialization: CPU +0, GPU +73, now: CPU 0, GPU 73 (MiB)
2024-05-24 12:57:30.353536943  [2024-05-24 12:57:30] frigate.detectors.plugins.tensorrt INFO    : [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +0, GPU +8, now: CPU 128, GPU 361 (MiB)
2024-05-24 12:57:30.353607367  [2024-05-24 12:57:30] frigate.detectors.plugins.tensorrt INFO    : [MemUsageChange] Init cuDNN: CPU +0, GPU +8, now: CPU 128, GPU 369 (MiB)
2024-05-24 12:57:30.353665257  [2024-05-24 12:57:30] frigate.detectors.plugins.tensorrt INFO    : [MemUsageChange] TensorRT-managed allocation in IExecutionContext creation: CPU +0, GPU +60, now: CPU 0, GPU 133 (MiB)

Operating system

Debian

Install method

Docker Compose

Coral version

CPU (no coral)

Any other information that may be helpful

local@Roberto:~$ docker exec -it frigate /bin/sh

nvidia-smi

Fri May 24 13:17:55 2024
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 525.147.05 Driver Version: 525.147.05 CUDA Version: 12.0 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|===============================+======================+======================|
| 0 NVIDIA GeForce ... On | 00000000:0B:00.0 Off | N/A |
| 31% 53C P2 51W / 175W | 570MiB / 8192MiB | 0% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+

+-----------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=============================================================================|
+-----------------------------------------------------------------------------+

d

[NickM-27]

The minimum driver version is 535

[KATPWR]

Thank you! Debian 12 repo is only up to 525. Added the Nvidia Developer Repo (instructions below) and got 545, now I'm good to go.

curl -fSsL https://developer.download.nvidia.com/compute/cuda/repos/debian12/x86_64/3bf863cc.pub | sudo gpg --dearmor | sudo tee /usr/share/keyrings/nvidia-drivers.gpg > /dev/null 2>&1

echo 'deb [signed-by=/usr/share/keyrings/nvidia-drivers.gpg] https://developer.download.nvidia.com/compute/cuda/repos/debian12/x86_64/ /' | sudo tee /etc/apt/sources.list.d/nvidia-drivers.list

sudo apt update && sudo apt upgrade