Tianyi Chen

[META] Tracking the code migration from ray repo

1

### Description #### Source code compilation and installation - [x] streaming c++ - [x] streaming python - [x] streaming java - [x] mobius python - [ ] mobius java ####...

Fix jave test.

1

Fix java test case for native issues.

### What changes were proposed in this pull request? 1. Implement ```CheckTrainingHangOperator``` based on XPU Timer metric. 2. Integrate context from ```JobManager``` in ```DiagnosisDataManager```. 3. Use limited ```Deque``` instead of...

[WIP] Optimize heartbeat timeout judgement if worker already succeeded.

1

### What changes were proposed in this pull request? 1. Add 'succeeded report' implement. 2. Add 'succeeded' flag for ```Node``` object. 3. Skip 'succeeded node' in 'noheartbeat' judgement. ### Why...

Enhance/Replace k8s python client.

2

# Background Currently, DLRover uses the official Kubernetes Python client to interact with the Kubernetes API Server. This part of implementations are quite important because it involves managing the lifecycle...

Introducing DLRover's solution for general AI computing[Proposal]

1

### What changes were proposed in this pull request? 1. A POC framework definition, along with a basic RL solution implementation. This includes the entire process from RL job submission...

**Is your feature request related to a problem? Please describe.** It is now recommended to directly use the checkpoint implementation from the latest version of Megatron. The dlrover's integration with...

Support HTTP for master-worker communication.

1

**Is your feature request related to a problem? Please describe.** Considering the use of gRPC involves significant dependency issues, and there is a degree of uncertainty when using gRPC in...

【WIP】Document updating for next release.

1

### What changes were proposed in this pull request? 1. Complete the documentation for the new architecture. 2. Update the homepage. 3. Update release-related information. ### Why are the changes...