Hello, First of all, thank you for all your responses. The

bentofile.yaml

was configured as follows. In the configuration, the

models

field is set to

"yolov8_optdemo:latest"

. After configuring this and running

bentoml build

, the

bento.yaml

file generated in the bento store contains the corresponding model information correctly. Additionally, the model is present in the model store as expected. However, the actual model folder from the model store is not copied (published) into the bento within the bento store. Why is this not happening? • As a reference, in this situation, if I manually copy the model folder from the model store into the bento within the bento store and deploy the bento to the production server, it runs properly using the model specified in the

bento.yaml

and operates without issues. ======================================================================== service: "app.service:AIServingService" include: - 'data/input' - 'app/*' - '.env' - '.env.testing' - '.env.development' - 'memo.txt' python: requirements_txt: requirements.txt models: - "yolov8_optdemo:latest" exclude: - "**/__pycache__" - "**/*.pyc" - "logs" - "data/output" - 'app/_test' ========================================================================= I have reviewed various possible issues, but the model is not being deployed. Did I configure something incorrectly? please check it.

Hello, I have missed this, what is the error that you face without running this manual model copy step?

PS D:\_gitlab\ai_framework\ai_serving_service> bentoml build -f bentofile_yolov8_optdemo.yaml ############################################# APP_ENV : development 2024-12-23 133302,181 - bentoml._internal.bento.build_config - INFO - Locking PyPI package versions. 2024-12-23 133302,182 - bentoml._internal.bento.build_config - INFO - Locking packages for x86_64-unknown-linux-gnu. Pass

--platform

option to specify the platform. ██████╗ ███████╗███╗ ██╗████████╗ ██████╗ ███╗ ███╗██╗ ██╔══██╗██╔════╝████╗ ██║╚══██╔══╝██╔═══██╗████╗ ████║██║ ██████╔╝█████╗ ██╔██╗ ██║ ██║ ██║ ██║██╔████╔██║██║ ██╔══██╗██╔══╝ ██║╚██╗██║ ██║ ██║ ██║██║╚██╔╝██║██║ ██████╔╝███████╗██║ ╚████║ ██║ ╚██████╔╝██║ ╚═╝ ██║███████╗ ╚═════╝ ╚══════╝╚═╝ ╚═══╝ ╚═╝ ╚═════╝ ╚═╝ ╚═╝╚══════╝ Successfully built Bento(tag="ai_serving_service:722uusoa42prlden"). Next steps: * Deploy to BentoCloud: $ bentoml deploy ai_serving_service:722uusoa42prlden -n ${DEPLOYMENT_NAME} * Update an existing deployment on BentoCloud: $ bentoml deployment update --bento ai_serving_service:722uusoa42prlden ${DEPLOYMENT_NAME} * Containerize your Bento with `bentoml containerize`: $ bentoml containerize ai_serving_service:722uusoa42prlden * Push to BentoCloud with `bentoml push`: $ bentoml push ai_serving_service:722uusoa42prlden PS D:\_gitlab\ai_framework\ai_serving_service> -------------------------------------------------------------------------------------------- There are no errors, and the build succeeds as shown above, but the model is not deployed. Here is the

bento.yaml

of the generated Bento. -------------------------------------------------------------------------------------------- service: app.service:AIServingService name: ai_serving_service version: 722uusoa42prlden bentoml_version: 1.3.14 creation_time: '2024-12-23T043303.165242+00:00' labels: {} models: - tag: yolov8_optdemo:iam3eg5mr2nkbden module: bentoml.pytorch creation_time: '2024-11-27T070723.255398+00:00' entry_service: ai_serving_service services: - name: ai_serving_service service: '' models: [] dependencies: - service: inference_service -------------------------------------------------------------------------------------------- Is there any example of deploying a model? The following documentation does not include such an example. https://docs.bentoml.com/en/latest/reference/bentoml/bento-build-options.html#models