All sources are now pushed to launchpad.net 
Give it anther hour or so until all the 4.23 packages are finally published.
After you got all the 4.23 packages installed and rebooted your RasPi5 (there is also a new firmware version), pull the hailo-rpi5-examples and make the following changes:
ubuntu@ubuntu-2404-pi5b:~/20251102_hailo-rpi5-examples$ git diff config.yaml
diff --git a/config.yaml b/config.yaml
index ead71dd..2a88c34 100644
--- a/config.yaml
+++ b/config.yaml
@@ -2,17 +2,17 @@
# This is RPI specefic variables
hailo_apps_infra_path: "auto" # Path to the Hailo Apps Infra repository, relative to the current working directory
-hailo_apps_infra_branch_tag: "25.7.0" # Branch or tag of the Hailo Apps Infra repository to use
+hailo_apps_infra_branch_tag: "dev" # Branch or tag of the Hailo Apps Infra repository to use
hailo_apps_infra_repo_url: https://github.com/hailo-ai/hailo-apps-infra.git
# This is Hailo Apps Infra variables
-hailort_version: "auto" # Options: "auto" (detect from device), "latest" (latest version), or specific version
-tappas_version: "auto" # Version of the Tappas model to use, set to "auto" to detect automatically
-model_zoo_version: "v2.14.0" # Version of the model zoo to use
+hailort_version: "4.23.0" # Options: "auto" (detect from device), "latest" (latest version), or specific version
+tappas_version: "5.1.0" # Version of the Tappas model to use, set to "auto" to detect automatically
+model_zoo_version: "v2.17.0" # Version of the model zoo to use
host_arch: "auto" # Options: "rpi" (Raspberry Pi), "x86" (desktop/server), "arm" (generic ARM board) or "auto" (detect from device)
hailo_arch: "auto" # Options: "hailo8", "hailo8l", or "auto" (detect from device)
resources_path: "resources" # Path to the resources directory, relative to the current working directory
virtual_env_name: "venv_hailo_rpi_examples" # Name of the virtual environment to be created
-server_url: "http://dev-public.hailo.ai/2025_01" # URL of the server to download models from , This is static
+server_url: "http://dev-public.hailo.ai/2025_10" # URL of the server to download models from , This is static
storage_dir: "hailo_temp_resources" # Directory to store the deb and whl files , this is the default if set to auot or empty
tappas_variant: "auto" # Options : "tappas" or " tappa-core" or "Auto"
-tappas_postproc_path: "auto" # Path to the post-processing script for Tappas, relative to the current working directory. If set to "auto", it will be detected automatically.
\ No newline at end of file
+tappas_postproc_path: "auto" # Path to the post-processing script for Tappas, relative to the current working directory. If set to "auto", it will be detected automatically.
After that, the install.sh should work and the
source setup_env.sh
python3 basic_pipelines/detection.py
should work as expected after that;-)
I hope, that helps a bit;-)
– Marco