Mikhail Kravets

Sorry to hear that , @221V 😞 I'd like to help but I totally forgot Rustlang. So, I have to ask a help of the community

Hi @rdavis120, Yes, your understanding is correct. The training algorithm retrieves one random item per training iteration. Regarding the error, it is zeroed during each backward step (see [`FeedForward::backward`](https://github.com/MikhailKravets/NeuroFlow/blob/c4df80b036b79aeb896e8ce7aab90eb8deb45f42/src/lib.rs#L370) method)....

@Tabrizian, ideally, it would be nice to have one library that automatically understands where it's run (or installed). So, `InferenceRequest`, `InferenceResponse`, and other objects from the lib will have one...

Logs of `kserve-container` when pod is initialized from zero by request ```shell 2023-05-06 15:09:20,049 [mlserver.parallel] DEBUG - Starting response processing loop... 2023-05-06 15:09:20,050 [mlserver.rest] INFO - HTTP server running on...

This error doesn't happen with `triton` inference server. Can it be the problem with `mlserver`'s `/v2/models//ready` endpoint? 🤔

Knative activator logs an error as well ```shell {"severity":"ERROR","timestamp":"2023-05-12T11:13:57.241782569Z","logger":"activator","caller":"net/revision_backends.go:398","message":"Failed to probe clusterIP 10.105.112.239:80","commit":"2c1bb07","knative.dev/controller":"activator","knative.dev/pod":"activator-6b9dc4c9db-kxkh9","knative.dev/key":"default/sklearn-iris-predictor-default-00001","error":"error roundtripping http://10.105.112.239:80/healthz: context deadline exceeded","stacktrace":"knative.dev/serving/pkg/activator/net. ``` Error itself ``` error roundtripping http://10.105.112.239:80/healthz: context deadline exceeded ```

Hi guys! Are there any updates on this?

Thanks for the answer @yetone. Is it possible to use AWS ECR outside of Amazon EKS then?

Should be fixed in #8. Please, feel free to re-open the issue if the error remains

Great. Thank you!!