Hello,
I’m currently working with the Metis Compute Board and the Voyager SDK, and I had a question regarding development outside of the SDK.
Is it possible to run a script like a custom inference or a similar script directly on the board without going through the Voyager SDK pipeline ?
I’m mainly asking to better understand the level of low-level access available on the Metis (runtime, APIs, or tools outside the SDK).
Thanks in advance for your help.
Best regards,
+2Ah, interesting approach! Let me ask around to get some opinions on that, and see what we can find out
Hello
The Voyager SDK virtual environment (venv) provides a comprehensive environment for running applications on Metis. It offers APIs at multiple abstraction levels, enabling users to integrate inference on Metis into their applications:
| Abstraction level | Low | Mid | High |
| API | AxRuntime | AxInferenceNet | InferenceStream |
| Scope | Neural network model | Model pipeline | End-to-end pipeline |
| App. definition | C++ and python | C++ (python following) | YAML |
| Reference Links | axruntime.md examples/axruntime | axinferencenet.md examples/axinferencenet | application.md |
On the Metis Compute Board, the Yocto OS based BSP makes direct package installation challenging. The recommended approach is to launch the Voyager SDK environment via a Docker container. From within this environment, you can select the appropriate APIs to run inference on Metis.
Could you share if there’s any specific constraint requiring you to run applications outside the Docker-based Voyager SDK environment? Understanding your needs will help us guide you further.
Already have an account? Login
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.
