docs: add Hy3 integrations guides (Part A) - #201
Open
smd-610 wants to merge 5 commits into
Open
Conversation
Update README files to include dataset preparation instructions and clarify key parameters for Hy3 reinforcement learning training. Enhance README files with acknowledgements for support from Tencent Hunyuan team and relevant communities. Fix typo in README by changing "Posttraining" to "Post-training" for consistency and clarity. Fix RL Post-training TOC anchor in README
docs: add RL post-training guide
There was a problem hiding this comment.
Pull request overview
This PR expands Hy3’s documentation by adding end-user integration guides for multiple AI tools under docs/integrations/, and also introduces reinforcement-learning (GRPO/verl) post-training documentation plus README entry points.
Changes:
- Add integrations landing page and per-tool setup guides (OpenCode Desktop, Claude Code CLI + VS Code plugin via Anthropic bridge, Cline, Kilo Code).
- Add new RL training guides (
rl/README*.md) and link them from the main READMEs. - Refine finetuning docs with a more precise note about which “bias not loaded” warnings are safe to ignore.
Reviewed changes
Copilot reviewed 12 out of 23 changed files in this pull request and generated 3 comments.
Show a summary per file
| File | Description |
|---|---|
docs/integrations/index.md |
Adds a consolidated integrations index with shared configuration info and tool list. |
docs/integrations/opencode-desktop.md |
Adds OpenCode Desktop setup guide for Hy3 via OpenAI-compatible endpoint. |
docs/integrations/claudecode.md |
Adds Claude Code CLI setup via Anthropic-compatible bridge guidance. |
docs/integrations/claudecode-plugin.md |
Adds Claude Code VS Code plugin setup via Anthropic-compatible bridge guidance. |
docs/integrations/cline.md |
Adds Cline setup guide for Hy3 via OpenAI-compatible endpoint. |
docs/integrations/kilocode.md |
Adds Kilo Code setup guide for Hy3 via OpenAI-compatible endpoint. |
rl/README.md |
Adds English GRPO/verl RL post-training guide and recommended environment/config notes. |
rl/README_CN.md |
Adds Simplified Chinese version of the RL post-training guide. |
README.md |
Adds “RL Post-training” section linking to the RL guide. |
README_CN.md |
Adds “强化学习训练” section linking to the RL guide. |
finetune/README.md |
Clarifies which bias-loading warnings are ignorable and which are not. |
finetune/README_CN.md |
Same clarification as above in Chinese. |
Suppressed comments (1)
docs/integrations/cline.md:39
- Step-by-step config uses
https://tokenhub.tencentmaas.comas Base URL, but later notes say the Base URL should end at/v1. Update this example to include/v1so users don't copy a potentially wrong URL.
1. 点齿轮进入 Settings。
2. Provider 下拉选择 **OpenAI Compatible**。
3. Base URL 填：`https://tokenhub.tencentmaas.com`
4. API Key 填：`HY3_API_KEY`。
5. Model ID 填：`hy3`。
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
Comment on lines
+16
to
+20
| | 协议 | OpenAI 兼容 | | ||
| | Provider | `OpenAI Compatible` | | ||
| | Base URL | `https://tokenhub.tencentmaas.com` | | ||
| | API Key | 你的 `HY3_API_KEY` | | ||
| | Model ID | `hy3` | |
Comment on lines
+55
to
+60
| - **OpenCode 桌面端**（OpenAI 兼容）：GUI 中 Provider 选 `OpenAI Compatible`，Base URL `https://tokenhub.tencentmaas.com/v1`、Key、Model(`hy3`) 填入即可。 | ||
| - **Claude Code**（Anthropic 协议）：需经 Anthropic 兼容桥接层（网关或 LiteLLM），再把 `ANTHROPIC_BASE_URL` / `ANTHROPIC_API_KEY` 指向它，模型映射为 `hy3`。 | ||
| - **Claude Code 插件**（Anthropic 协议）：同 Claude Code，但环境变量为 `ANTHROPIC_BASE_URL` / `ANTHROPIC_AUTH_TOKEN` / `ANTHROPIC_MODEL`，模型 `hy3`。 | ||
| - **Cline**（OpenAI 兼容）：Provider 选 `OpenAI Compatible`，Base URL `https://tokenhub.tencentmaas.com`、Key、Model ID(`hy3`)。 | ||
| - **Kilo Code**（OpenAI 兼容）：API 提供商选 `OpenAI 兼容`，基础 URL `https://tokenhub.tencentmaas.com/v1`、密钥、模型(`hy3`)。 | ||
|
|
Comment on lines
+209
to
+212
| ## RL Post-training | ||
|
|
||
| Hy3 supports GRPO reinforcement learning training with [verl](https://github.com/volcengine/verl), training on Megatron-LM (model conversion via NVIDIA Megatron-Bridge) with vLLM rollout. For detailed documentation, please refer to: [RL Training Guide](./rl/README.md) | ||
|
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
PR：Hy3 主流 AI 产品接入指南（Part A）+ 小作品（Part B）
认领信息
rhinobird2026Part A：Hy3 主流 AI 产品接入指南（本 PR）
覆盖 5 个工具的完整接入指南，放入
docs/integrations/：每篇均含：安装与版本要求 → 配置项（Base URL / Model / 鉴权 / 协议）→ 第一次对话 → 跑通真实任务（新建并运行
hello.py）+ 常见注意事项。docs/integrations/index.md，含通用接入信息与curl自检脚本。docs/integrations/screenshots/（每个工具 配置 + 跑通任务 两张）。Part B：Hy3 小作品《Hy3 幻境跑团》
tool_calls真实执行（掷骰 / 扣血 / 发物品）reasoning_content）README.md+demo-video.mp4（≤1min）+ 完整源码（API Key 仅存服务端，已 gitignore）