Skip to content

Comments

[doc] ascend add performance analysis guide and update some version info#5324

Open
chengminhua wants to merge 5 commits intoverl-project:mainfrom
chengminhua:main
Open

[doc] ascend add performance analysis guide and update some version info#5324
chengminhua wants to merge 5 commits intoverl-project:mainfrom
chengminhua:main

Conversation

@chengminhua
Copy link
Contributor

@chengminhua chengminhua commented Feb 15, 2026

What does this PR do?

ascend add performance analysis guide and update some version info

Checklist Before Starting

  • Search for similar PRs. Paste at least one query link here: ...
  • Format the PR title as [{modules}] {type}: {description} (This will be checked by the CI)
    • {modules} include fsdp, megatron, veomni, sglang, vllm, rollout, trainer, ci, training_utils, recipe, hardware, deployment, ray, worker, single_controller, misc, perf, model, algo, env, tool, ckpt, doc, data, cfg, reward
    • If this PR involves multiple modules, separate them with , like [megatron, fsdp, doc]
    • {type} is in feat, fix, refactor, chore, test
    • If this PR breaks any API (CLI arguments, config, function signature, etc.), add [BREAKING] to the beginning of the title.
    • Example: [BREAKING][fsdp, megatron] feat: dynamic batching

Test

For changes that can not be tested by CI (e.g., algorithm implementation, new model support), validate by experiment(s) and show results like training curve plots, evaluation results, etc.

API and Usage Example

Demonstrate how the API changes if any, and provide usage example(s) if possible.

# Add code snippet or script demonstrating how to use this

Design & Code Changes

Demonstrate the high-level design if this PR is complex, and list the specific changes.

Checklist Before Submitting

Important

Please check all the following items before requesting a review, otherwise the reviewer might deprioritize this PR for review.

@chengminhua chengminhua changed the title add some file [doc] ascend add performance analysis guide and update some version info Feb 15, 2026
Copy link
Contributor

@gemini-code-assist gemini-code-assist bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Code Review

此 PR 为昇腾 NPU 环境引入了性能分析指南,并同步更新了快速入门文档和 Docker 构建指南。主要改进包括修正了 Dockerfile 构建路径以及明确了 transformers 的版本依赖。性能分析指南内容详实,对开发者具有很高的参考价值。目前发现的主要问题是性能分析指南中的图片链接使用了 GitHub 的 blob 路径,这会导致图片无法在 Markdown 页面中正常显示,建议统一更换为 raw 链接。

4. **reward**:基于prompt和生成的response,奖励模型评估奖励值R_N。
5. **update**:基于计算得到的R_N、ref logprob、logprob计算优化函数和策略梯度,对actor模型进行更新

![rl_data_stream](https://github.com/chengminhua/verl_data/blob/main/MindStudio_Insight_use/rl_data_stream.png)
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

high

GitHub 的 blob 链接在 Markdown 中无法直接渲染为图片。建议将链接中的 blob 替换为 raw,以确保图片能正常显示。此问题在以下行中也存在,请一并修改:31, 40, 49, 61, 71, 79, 89, 93, 107, 111, 121。

Suggested change
![rl_data_stream](https://github.com/chengminhua/verl_data/blob/main/MindStudio_Insight_use/rl_data_stream.png)
![rl_data_stream](https://github.com/chengminhua/verl_data/raw/main/MindStudio_Insight_use/rl_data_stream.png)

@chengminhua
Copy link
Contributor Author

@wucong25 brother cong,please help me review.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant