feat(training):support online training

[qingTang0305]

Change-Id: Ia388f09600ef9547ab2fe16d595ff01a4ed4dca1

AgentScope-Java Version

[The version of AgentScope-Java you are working on, e.g. 1.0.8, check your pom.xml dependency version or run mvn dependency:tree | grep agentscope-parent:pom(only mac/linux)]

Description

feat(training):#702

Checklist

Please check the following items before code is ready to be reviewed.

• Code has been formatted with mvn spotless:apply
• All tests are passing (mvn test)
• Javadoc comments are complete and follow project conventions
• Related documentation has been updated (e.g. links, examples, etc.)
• Code is ready for review

[gemini-code-assist]

Summary of Changes

Hello @qingTang0305, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request introduces a robust online training extension for AgentScope Java, designed to facilitate continuous improvement of agents by integrating with the Trinity-RFT framework. It establishes a data feedback loop from production to training systems, enabling agents to learn and adapt from real-time interactions. The extension provides automated mechanisms for data collection, reward calculation, and training request management, aiming for a minimal setup experience for developers.

Highlights

• New Training Extension Module: A new Maven module, agentscope-extensions-training, has been added to enable online training capabilities for AgentScope Java applications.
• Trinity-RFT Integration: Introduces TrinityClient for seamless interaction with the Trinity-RFT backend, supporting feedback and commit APIs for continuous model optimization.
• Automated Task/Run ID Management: The module provides automated generation and management of Task IDs and Run IDs through TaskIdGenerator, RunRegistry, and RunExecutionContext, simplifying training data tracking.
• Shadow Traffic Training Architecture: A TrainingRouter hook implements a shadow traffic training architecture, allowing a sample portion of production requests to be processed by a 'shadow agent' for training without impacting live service.
• Flexible Training Selection Strategies: Two built-in strategies, SamplingRateStrategy and ExplicitMarkingStrategy, are provided for filtering requests for training, along with an interface for custom selection logic.
• Customizable Reward Calculation: A RewardCalculator interface allows developers to define custom logic for calculating reward values based on shadow agent execution results, crucial for reinforcement learning.
• Comprehensive Documentation: Detailed documentation in both English and Chinese has been added, covering the overview, architecture, core features, quick start guide, and configuration options for the online training extension.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	/gemini review	Performs a code review for the current pull request in its current state.
Pull Request Summary	/gemini summary	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	/gemini help	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

[gemini-code-assist]

Code Review

This pull request introduces a comprehensive online training extension, which is a significant and well-structured feature. The code is generally of high quality, with good use of reactive programming and clear separation of concerns. I've identified a few areas for improvement, primarily around error handling and documentation consistency, to make the new module more robust and easier to use.

[codecov]

Codecov Report

❌ Patch coverage is 77.00394% with 175 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
...gentscope/core/training/runner/TrainingRouter.java	36.80%	77 Missing and 2 partials ⚠️
...o/agentscope/core/training/runner/AgentCloner.java	0.00%	52 Missing ⚠️
...ope/core/training/backend/TrinityModelAdapter.java	71.05%	11 Missing ⚠️
...ope/core/training/strategy/TrainingAnnotation.java	77.77%	0 Missing and 10 partials ⚠️
...gentscope/core/training/runner/TrainingRunner.java	92.55%	6 Missing and 1 partial ⚠️
...cope/core/training/runner/RunExecutionContext.java	92.68%	0 Missing and 3 partials ⚠️
...pe/core/training/runner/TaskExecutionRegistry.java	95.23%	0 Missing and 3 partials ⚠️
...gentscope/core/training/runner/TrainingConfig.java	95.83%	2 Missing and 1 partial ⚠️
...ore/training/strategy/ExplicitMarkingStrategy.java	91.66%	1 Missing and 2 partials ⚠️
...e/training/strategy/TrainingSelectionStrategy.java	0.00%	2 Missing ⚠️
... and 2 more

📢 Thoughts on this report? Let us know!