feat(agentic): add GraphGPO ALFWorld recipe - #280
Draft
huxy1225-cloud wants to merge 1 commit into
Draft
Conversation
Contributor
|
你好,可以进一步精简一下代码结构、文件组织,使其达到可 review 状态,再补充更多的实验内容(复现指令、关键 metrics 图等) |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
改动内容
我在 Relax 中加入了 ALFWorld GraphGPO recipe,并补充了多轮轨迹使用 variable-row 和逐
turn 自定义 advantage 所需的框架支持。
主要改动包括:
variable-row 元数据;
未启用 GraphGPO 时继续使用原有训练路径。
Related to #225.
验证情况
冻结候选的历史验证记录:
我已把改动合并到提交时最新的
main(682d474e2b22325d5f63706af35f644e7a21ccc3)。当前本机没有项目测试环境,因此没有在这个合并后的 commit 上重新运行 pytest 或
pre-commit;CI 结果以 PR 页面为准。
当前复现结果
目前保留了 5 条完整的 150-step、eval149、128-episode 最终评估:
这两组配置不能组成严格的 12-run 横向矩阵。当前结果只能作为实现和部分复现证据,不能
声称已经完成 Task37 论文复现或通过验收。后续仍需补齐同一冻结配置下的 4 种方法 × 3
个 seed、完整测试输出和总训练时间开销比例。
变更类型