Li Jiang

[Feature Request]: Need a navigation bar or site map

1

### Is your feature request related to a problem? Please describe. Can't find a navigation bar or site map to help jump from one page to another.  ### Describe...

## Why are these changes needed? Fix typos, upgrade yarn packages, add some improvements ## Related issue number ## Checks - [ ] I've included any doc changes needed for...

[Issue]: Why mypy check is skipped in merge checks?

5

### Describe the issue For instance, in PR #2466, mypy check should not be skipped. https://github.com/microsoft/autogen/actions/runs/8772891100/job/24072334718  ### Steps to reproduce _No response_ ### Screenshots and logs _No response_ ###...

Is there a better way to execute the generated code?

12

The current code execute logic has an issue with code generated in several blocks and later blocks depends on the former ones. For example: > User_Proxy (to Assistant_Agent): > >...

joblib-spark doesn't work with joblib 1.4.0

1

joblib 1.4.0 just released two days ago, it will break joblib-spark: ``` if SafeFunction is None: # pylint: disable=protected-access,used-before-assignment return self._get_pool().apply_async( > PoolManagerMixin._wrap_func_call, (run_on_worker_and_fetch_result,), callback=callback, error_callback=callback ) E AttributeError: type...

ERNIE模型对message的轮次，role的限制

3

如下图所示，错误示例及[错误代码](https://cloud.baidu.com/doc/WENXINWORKSHOP/s/tlmyncueh)。ERNIE模型对message的轮次，role有一定的限制，但这些限制会导致模型无法在复杂场景下应用。   这个限制在openAI里是没有的： 

### Is your feature request related to a problem? Please describe. Sometimes I just want to silent a specific agent in the chats. ### Describe the solution you'd like Add...

## Why are these changes needed? Current RAG agents RetrieveUserProxyAgent and RetrieveAssistantAgent work together to provide RAG functionality. However, using one single RAG agent would be more convenient for group...

It would be useful to support parallel tunning in openAI integration to accelerate the tunning process. Plus, it would be better to support multi-apikeys to mitigate the API call ratio...