Check CPU AVX instruction support before installing OpenVino model (#2796)

This commit is contained in:
Rémi Bédard-Couture 2024-04-16 05:39:00 -04:00 committed by GitHub
parent 9cfad87e3b
commit 049a8096d3
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 33 additions and 20 deletions

View File

@ -77,18 +77,6 @@ sed -i '/^s6-svc -O \.$/s/^/#/' /opt/frigate/docker/main/rootfs/etc/s6-overlay/s
cat <<EOF >/config/config.yml
mqtt:
enabled: false
detectors:
ov:
type: openvino
device: AUTO
model:
path: /openvino-model/FP16/ssdlite_mobilenet_v2.xml
model:
width: 300
height: 300
input_tensor: nhwc
input_pixel_format: bgr
labelmap_path: /openvino-model/coco_91cl_bkgr.txt
cameras:
test:
ffmpeg:
@ -109,13 +97,38 @@ sed -i -e 's/^kvm:x:104:$/render:x:104:root,frigate/' -e 's/^render:x:105:root$/
echo "tmpfs /tmp/cache tmpfs defaults 0 0" >> /etc/fstab
msg_ok "Installed Frigate $RELEASE"
msg_info "Installing Object Detection Models (Resilience)"
if grep -q -o -m1 'avx[^ ]*' /proc/cpuinfo; then
echo -e "AVX support detected"
msg_info "Installing Openvino Object Detection Model (Resilience)"
$STD pip install -r /opt/frigate/docker/main/requirements-ov.txt
cd /opt/frigate/models
export ENABLE_ANALYTICS=NO
$STD /usr/local/bin/omz_downloader --name ssdlite_mobilenet_v2 --num_attempts 2
$STD /usr/local/bin/omz_converter --name ssdlite_mobilenet_v2 --precision FP16 --mo /usr/local/bin/mo
cd ..
cat <<EOF >>/config/config.yml
detectors:
ov:
type: openvino
device: AUTO
model:
path: /openvino-model/FP16/ssdlite_mobilenet_v2.xml
model:
width: 300
height: 300
input_tensor: nhwc
input_pixel_format: bgr
labelmap_path: /openvino-model/coco_91cl_bkgr.txt
EOF
msg_ok "Installed Openvino Object Detection Model (Resilience)"
else
cat <<EOF >>/config/config.yml
model:
path: /cpu_model.tflite
EOF
fi
msg_info "Installing Coral Object Detection Model (Resilience)"
cd /opt/frigate
export CCACHE_DIR=/root/.ccache
export CCACHE_MAXSIZE=2G
wget -q https://github.com/libusb/libusb/archive/v1.0.26.zip
@ -142,7 +155,7 @@ wget -qO cpu_audio_model.tflite https://tfhub.dev/google/lite-model/yamnet/class
cp /opt/frigate/audio-labelmap.txt /audio-labelmap.txt
mkdir -p /media/frigate
wget -qO /media/frigate/person-bicycle-car-detection.mp4 https://github.com/intel-iot-devkit/sample-videos/raw/master/person-bicycle-car-detection.mp4
msg_ok "Installed Object Detection Models"
msg_ok "Installed Coral Object Detection Model"
msg_info "Building Nginx with Custom Modules"
$STD /opt/frigate/docker/main/build_nginx.sh