Richard Liu

Kubeflow component integration with ML Metadata

9

/kind feature **Why you need this feature:** Kubeflow currently doesn't have a unified metadata/artifact management story beyond what's supported in KFP. For example, the concept of a "ML experiment" exists...

[Feature] Support batch scheduling and queueing

4

### Search before asking - [X] I had searched in the [issues](https://github.com/ray-project/kuberay/issues) and found no similar feature requirement. ### Description Kuberay currently does not seem to support scheduling policies for...

Are there examples compatible with TF 2.0, similar to https://github.com/tensorflow/ecosystem/tree/master/distribution_strategy?

[Feature] Integrate KubeRay with Kubeflow

1

### Search before asking - [X] I had searched in the [issues](https://github.com/ray-project/kuberay/issues) and found no similar feature requirement. ### Description KubeRay can be installed with other OSS ML Ops products...

## Why are these changes needed? ## Related issue number ## Checks - [ ] I've made sure the tests are passing. - Testing Strategy - [ ] Unit tests...

[Feature] Add support for TPUs in Kuberay operator

1

### Search before asking - [X] I had searched in the [issues](https://github.com/ray-project/kuberay/issues) and found no similar feature requirement. ### Description GKE will add support for TPU VMs in early 2023....

Integrate katib with Jupyter notebooks and fairing

5

Allow data scientists to take a training job on fairing and easily configure a katib HP tuning job - without having to write a CRD.

Refactoring common operator tests

5

There are common Python test code in: https://github.com/kubeflow/pytorch-operator/tree/master/test https://github.com/kubeflow/tf-operator/tree/master/test Consider how these can be refactored to reduce duplication.

### Search before asking - [X] I had searched in the [issues](https://github.com/ray-project/kuberay/issues) and found no similar feature requirement. ### Description Add a "GKE provider" to the Ray Cluster launcher that...

Empty response returned for prompt responses when using run_server_with_ray.py and batch_size > 1

2

Sending multiple prompts to the server, only the first prompt is able to return any results. Requests after the first one would only return an empty response. I've tried 3...