Aaron Pham

Can you update the buildx version to the latest one and try again?

This can be done by reading https://github.com/tiangolo/fastapi/blob/5c576e42be8f8832f0e225beaa72153a256892d7/fastapi/applications.py#L197 and merge with BentoML [generated OpenAPI spec](https://github.com/bentoml/BentoML/blob/a48d83e07f04a881c4550ca652d1e80309fbbbf4/bentoml/_internal/service/service.py#L196) Would love for community to contribute on this feature 🎉

Hi @ramtiin, Can you share the whole training code for the learner?

Feel free to reopen if you are still running into this.

I was thinking about ```python @svc.api(input=Union[JSON, NumpyNdarray], output=Union[]) def endpoint(input): return JSON.from_sample({"my_field": my_value}), NumpyNdarray.from_sample(np.array([[1,2,3,]]) ```

Hi there, one small request. Is it possible that you can move the `client` folder to `bentoml` instead of `bentoml._internal`? I think it is nice to also expose the python...

> > Notice the input / output option of local / remote (s3 etc...) files. We are currently using PyFilesystem2, would be great if we can use that. to avoid...

I'm thinking to pass it as a ``config-options`` ```bash bentoml serve --production --config-options 'runners.pytorch_mnist.resources."nvidia.com/gpu"=[0,2]' --config-options "api_server.http.metrics.enabled=false" ```

I'm thinking to just run all general CI jobs nonetheless

Eh, I will close this for now since 3.11 is not fully supported by many of our dependent libraries. We will run this back when 3.11 is out officially. I...