Configuration¶
Find the BentoML helper commands, API ports, and access-file guidance for your instance.
Where things live¶
| Item | What it is |
|---|---|
| Port 3000 | Swagger UI, after you run sushi-bentoml-example. |
| Port 8448 | Prediction API. A per-instance token protects it. |
sushi-bentoml-api |
Helper that reads the access file. The listing does not name the file path. |
sushi-bentoml-example |
Starts the browser example. |
/opt/sushi/docs/demo.sh |
Walkthrough. |
/opt/sushi/docs/README.md |
Helper arguments and the access file. |
| Sample model | Bundled scikit-learn model for the first prediction. |
Two ways to call it¶
- Browser:
sushi-bentoml-example, forward port 3000, then Try it out and Execute in Swagger UI. - API: forward port 8448 and use
sushi-bentoml-api, which reads the token for you.
Do not print the token. Do not put it on the command line.
Listen address¶
Leave inbound application ports closed. Both ports are reached through Session Manager. See Open Swagger and Call the prediction API.
Last updated November 1, 2026 · Suggest an improvement