Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Python: Fix agent group chat bug related to function calling in ChatCompletionAgent #8330

Merged
merged 2 commits into from
Aug 22, 2024

Conversation

moonbox3
Copy link
Contributor

Motivation and Context

During a group chat, if using an OpenAIAssistantAgent, the addition of a ChatCompletionMessage to the AssistantAgent channel was breaking because we were including FunctionCallContent. Additionally, there needs to be some translation of a FunctionResultContent type to a text type when adding a ChatCompletionAgent message to an AssistantAgent thread message. The AuthorRole can only be 'user' or 'assistant'.

Description

Fixes for the items mentioned above. This may not be a perfect solution, but getting this in and I can revisit it to align more closely with how dotnet handles the FunctionResultContent/TextContent as part of the overall ChatMessageContent.

  • This PR also adds a new sample showing mixed agent chat and a ChatCompletionAgent using plugins.

Contribution Checklist

@moonbox3 moonbox3 requested a review from a team as a code owner August 22, 2024 15:31
@markwallace-microsoft markwallace-microsoft added the python Pull requests for the Python Semantic Kernel label Aug 22, 2024
@markwallace-microsoft
Copy link
Member

markwallace-microsoft commented Aug 22, 2024

Python Test Coverage

Python Test Coverage Report
FileStmtsMissCoverMissing
semantic_kernel
   kernel.py1994776%150, 161, 165, 315–318, 425, 439–482
semantic_kernel/agents/channels
   open_ai_assistant_channel.py49198%42
semantic_kernel/agents/group_chat
   agent_chat.py116199%91
   agent_group_chat.py67297%113, 135
   broadcast_queue.py72199%35
semantic_kernel/agents/open_ai
   assistant_content_generation.py68297%81–82
   open_ai_assistant_base.py361399%243, 321–322
semantic_kernel/connectors/ai/anthropic/services
   anthropic_chat_completion.py942573%171–181, 196–205, 222–236
semantic_kernel/connectors/ai/azure_ai_inference/services
   azure_ai_inference_chat_completion.py150497%174, 324–325, 441
semantic_kernel/connectors/ai/embeddings
   embedding_generator_base.py8188%50
semantic_kernel/connectors/ai/google
   shared_utils.py39295%97, 119
semantic_kernel/connectors/ai/google/google_ai/services
   google_ai_chat_completion.py150497%153, 265, 299–300
   utils.py63395%139, 157–162
semantic_kernel/connectors/ai/google/vertex_ai/services
   utils.py65395%141, 159–164
   vertex_ai_chat_completion.py150497%147, 258, 292–293
semantic_kernel/connectors/ai/hugging_face/services
   hf_text_completion.py54394%97, 105, 136
   hf_text_embedding.py32584%79–83
semantic_kernel/connectors/memory/azure_ai_search
   azure_ai_search_collection.py87298%150, 152
semantic_kernel/connectors/memory/redis
   redis_collection.py160299%146, 316
   utils.py451176%145–146, 164, 166, 173–188
semantic_kernel/connectors/openapi_plugin
   openapi_manager.py58297%110–111
   openapi_parser.py88199%127
   openapi_runner.py105298%181–182
semantic_kernel/connectors/openapi_plugin/models
   rest_api_operation.py129199%242
semantic_kernel/contents
   function_call_content.py97199%201
semantic_kernel/data
   vector_store_record_collection.py2491992%410, 470–474, 482–486, 526–529, 536–539
   vector_store_record_utils.py26292%50, 52
semantic_kernel/functions
   kernel_function_decorator.py98199%102
   kernel_function_from_method.py96199%153
   kernel_function_from_prompt.py154795%165–166, 180, 201, 219, 239, 322
   kernel_function_log_messages.py33294%36–37
   kernel_plugin.py187299%472, 475
semantic_kernel/planners
   plan.py2344581%54, 163–165, 197, 214–227, 264, 269, 277–278, 288–291, 308, 313, 329, 332–337, 355, 360, 363, 365, 372, 386–388, 393–397
semantic_kernel/planners/function_calling_stepwise_planner
   function_calling_stepwise_planner.py116497%145, 189–190, 198
semantic_kernel/planners/sequential_planner
   sequential_planner.py64691%71, 75, 109, 125, 134–135
   sequential_planner_extensions.py50982%31–32, 56, 110–124
   sequential_planner_parser.py771284%66–74, 93, 117–120
semantic_kernel/schema
   kernel_json_schema_builder.py119596%50, 84, 206, 210–211
semantic_kernel/template_engine/blocks
   code_block.py77199%119
   named_arg_block.py43198%98
semantic_kernel/utils/telemetry
   decorators.py113496%232, 240, 243, 252
   user_agent.py16288%18–19
TOTAL1076725698% 

Python Unit Test Overview

Tests Skipped Failures Errors Time
2370 1 💤 0 ❌ 0 🔥 52.072s ⏱️

@moonbox3 moonbox3 added this pull request to the merge queue Aug 22, 2024
Merged via the queue into microsoft:main with commit 3a5d601 Aug 22, 2024
25 checks passed
@moonbox3 moonbox3 deleted the group_chat_fc branch August 22, 2024 16:21
LudoCorporateShark pushed a commit to LudoCorporateShark/semantic-kernel that referenced this pull request Aug 25, 2024
…ompletionAgent (microsoft#8330)

### Motivation and Context

During a group chat, if using an OpenAIAssistantAgent, the addition of a
ChatCompletionMessage to the AssistantAgent channel was breaking because
we were including FunctionCallContent. Additionally, there needs to be
some translation of a FunctionResultContent type to a text type when
adding a ChatCompletionAgent message to an AssistantAgent thread
message. The AuthorRole can only be 'user' or 'assistant'.

<!-- Thank you for your contribution to the semantic-kernel repo!
Please help reviewers and future users, providing the following
information:
  1. Why is this change required?
  2. What problem does it solve?
  3. What scenario does it contribute to?
  4. If it fixes an open issue, please link to the issue here.
-->

### Description

Fixes for the items mentioned above. This may not be a perfect solution,
but getting this in and I can revisit it to align more closely with how
dotnet handles the FunctionResultContent/TextContent as part of the
overall ChatMessageContent.
- This PR also adds a new sample showing mixed agent chat and a
ChatCompletionAgent using plugins.

<!-- Describe your changes, the overall approach, the underlying design.
These notes will help understanding how your code works. Thanks! -->

### Contribution Checklist

<!-- Before submitting this PR, please make sure: -->

- [X] The code builds clean without any errors or warnings
- [X] The PR follows the [SK Contribution
Guidelines](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md)
and the [pre-submission formatting
script](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md#development-scripts)
raises no violations
- [X] All unit tests pass, and I have added new tests where possible
- [X] I didn't break anyone 😄
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
python Pull requests for the Python Semantic Kernel
Projects
Status: Sprint: Done
Development

Successfully merging this pull request may close these issues.

5 participants