Skip to content

Releases: OpenCSGs/csghub-server

v1.14.0-ce

08 Dec 01:23
7064e9a

Choose a tag to compare

v1.14.0-ce

✨ New Features

  • Deployment Workflow Migration: Refactored the core Deploy Scheduler and successfully migrated its logic to use Temporal Workflow for improved reliability and scalability. ([#495])
  • Finetuning Job Support: Added support to submit fine-tune tasks directly as standard jobs. ([#522])
  • AI Gateway Integration: The AI Gateway now supports the MCP (Multi-Cluster Platform) for enhanced model serving capabilities. ([#529])
  • Jupyter Notebook Support: Application space deployment now supports Jupyter Notebook environments.
  • Internal Notification Tool: Introduced a new tool to generate basic internal notification messages. ([#551])

🚀 Enhancements & Bug Fixes

  • Dependency Updates: Bumped argo-workflows from 3.5.13 to 3.6.12. Upgraded casdoor SDK to v1.22.0 and updated user info retrieval to use UUID. ([#481], [#534])
  • API & Core: Enhanced middleware robustness. Ensured header keys are handled as case-sensitive. Enhanced concurrency in CheckRepoFiles and improved test coverage. Enabled soft delete functionality for licenses. ([#547], [#482], [#530], [#553])
  • Deployment & Runner: Improved runner network discovery logic to read clusterIP if no ingress IP is found. Added StorageClass support in cluster report events. ([#484], [#525])
  • Finetuning: Added model/dataset fields to finetune job operations for richer metadata. ([#527])
  • Integration: Added User-Email to request headers for DataFlow and Label Studio integrations. ([#492])
  • Git/LFS Fixes: Fixed a client bug preventing the deletion of remote branches. Fixed a cache bug in LFS sync upload ID. ([#521], [#489])
  • Temporal Workflow Fixes: Fixed command errors and a worker panic when the sensitive checker was not enabled. Added initialization sensitive checker in the temporal worker. ([#501], [#506], [#519])
  • Multi-Sync & Rproxy Fixes: Corrected the logic for setting the service host name in the rproxy header. Fixed a bug where multi-sync MCP servers failed to display file lists and READMEs. ([#485], [#533])
  • General Bug Fixes: Fixed bugs related to getting evaluations with access control, AI Gateway external LLM support, and repository creation. ([#529], [#532], [#536])

Full Changelog: v1.12.0-ce...v1.14.0-ce

v1.12.1-ce

19 Nov 02:28

Choose a tag to compare

v1.12.1-ce Pre-release
Pre-release

What's Changed

  • Add initialization for sensitive checkerfix fix init sensitive checke… by @phantom-rabbit in #507
  • Resolving sensitive check disable causes worker startup panic. by @CementZhang in #520

Full Changelog: v1.12.0-ce...v1.12.1-ce

v1.12.0-ce

05 Nov 07:50

Choose a tag to compare

New Features

  • Atomic Repository Creation: We've implemented atomic creation of repositories. This enhancement ensures that the $\text{git}$ operation and the database persistence complete as a single, indivisible unit. This prevents interruptions or inconsistencies that could previously occur between the file system operation and the database record, leading to a more robust and reliable system.

  • Automatic Runner Discovery and Cluster Auto-Scaling: The system now supports automatic discovery for remote runners. The $\text{csghub-server}$ will automatically detect and register runner services as they come online or go offline. This functionality enables cluster-level auto-scaling, allowing the platform to dynamically adjust resources based on demand.

  • Streamable Protocol for $\text{mcp}$ Space Execution: We have introduced a streamable protocol to execute $\text{mcp}$ (Model/Compute/Processing) spaces. This new protocol replaces Server-Sent Events ($\text{SSE}$) to deliver better performance and significantly improve compatibility with reverse proxies.

Enhancement

  • Tracing: enhance request handler with robust tracing, logging, and retry
  • MultiSync: allow filter local only repositories
  • Bug Fix: invalid var name in header for proxy csghub-dataflow request in k8s ingress

What's Changed

Full Changelog: v1.11.0-ce...v1.12.0-ce

v1.11.0-ce

12 Sep 01:37
d2cb30c

Choose a tag to compare

✨ New Features

  • Runner Service Refactored: The Runner service has been completely refactored. It now runs seamlessly both inside and outside a Kubernetes cluster and automatically reports its endpoint and configuration data to the CSGHub server. The Kubernetes cluster is now pluggable at runtime. Most importantly, if you configure a service account for the Runner service in Kubernetes, you no longer need a kubeconfig file, which is a significant security improvement.
  • Go Error Documentation Tool: We've added a new command-line tool that scans Go code for custom error comments and automatically generates error documentation. This tool also includes numerous pre-defined errors and multilingual translations.
  • Internationalized Notifications: Notification messages now support i18n (internationalization).

🚀 Enhancements & Bug Fixes

  • Distributed Lock for DB Migration: Added a distributed lock to prevent conflicts during database migrations.
  • Git Callback Fix: Fixed a bug where the Git callback was not triggered if a repository contained no LFS files.
  • Localized Git Errors: Error messages for Git failures are now localized.
  • Temporal Workflow Context Fix: Resolved an issue where an incorrect context was causing the Temporal workflow to terminate unexpectedly.
  • MCP Repository Mirroring: It's now possible to mirror and sync MCP repositories.
  • New File Deletion API: A new API has been added to delete files from repositories.

What's Changed

Full Changelog: v1.10.0-ce...v1.11.0-ce

v1.10.0-ce

16 Aug 10:16

Choose a tag to compare

New Features

  • DataFlow Tool: Introduced a one-click data processing feature with the new DataFlow tool. This tool can read datasets from CSGHub and save processed datasets back to CSGHub. For more details, please refer to the csghub-dataflow repository.

Enhancements

  • Model Inference: Improved model inference capabilities to support the ERNIE and Huanyuan models, as well as the ability to perform score model inference.

  • Sync Code Refactoring: Refactored and enhanced the synchronization code to allow users to cancel model or dataset syncing tasks.

  • AI Gateway Updates: Modified the response data structure in the AI Gateway to ensure compatibility with the OpenAI API when no running model is found.

  • Internationalization (i18n) Support: Expanded i18n support; all server-side APIs now offer customizable error messages that can be translated into three languages: Simplified Chinese (zh-CN), American English (en-US), and Traditional Chinese (zh-HK).

What's Changed

Full Changelog: v1.9.0-ce...v1.10.0-ce

v1.9.5

08 Aug 04:34
e4d7f59

Choose a tag to compare

What's Changed

Full Changelog: v1.9.4-ce...v1.9.5-ce

v1.9.4

05 Aug 05:53
86db48f

Choose a tag to compare

What's Changed

Full Changelog: v1.9.3-ce...v1.9.4-ce

v1.9.3

29 Jul 02:27
81dcb63

Choose a tag to compare

What's Changed

Full Changelog: v1.9.2-ce...v1.9.3-ce

v1.9.2

24 Jul 08:24
77436f1

Choose a tag to compare

What's Changed

Full Changelog: v1.9.1-ce...v1.9.2-ce

v1.9.1-ce

22 Jul 10:17
5e5cdac

Choose a tag to compare

What's Changed

Full Changelog: v1.9.0-ce...v1.9.1-ce