fix(response_format): changed default serialization

[giuliastf]

Description

• Updated the chat_completions method in llm_gateway_service.py to accept both lists of dictionaries and lists of tuples as message input.

• Improved the _simple_serialize_defaults function in tracing/_utils.py to handle Pydantic BaseModel instances and classes, including schema serialization for classes and preventing serialization of class types as instances.
  Fix Issue Invoking chat_completions with response_format=pydantic_model throws exception #635

[Copilot]

Pull Request Overview

This PR fixes response format handling and serialization issues in the UiPath Python SDK. The changes improve message format compatibility and resolve JSON serialization errors when using Pydantic BaseModel classes as parameters.

• Enhanced message format support to accept both dictionaries and tuples with role normalization
• Improved serialization handling for Pydantic BaseModel classes and instances
• Added comprehensive test coverage for the new functionality

Reviewed Changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 2 comments.

File	Description
src/uipath/_services/llm_gateway_service.py	Added role mapping and message format conversion to support tuples and normalize roles
src/uipath/tracing/_utils.py	Enhanced serialization to handle Pydantic BaseModel classes with schema representation
tests/tracing/test_traced.py	Added comprehensive test for Pydantic BaseModel serialization and tuple message handling
pyproject.toml	Version bump to 2.1.73

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[Copilot]

Pull Request Overview

Copilot reviewed 4 out of 4 changed files in this pull request and generated 3 comments.

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[akshaylive]

Thanks for addressing this! :)

[akshaylive]

LGTM 🚀

[radu-mocanu]

do we have a clear error message from llm gateway if the 'roles' passed are not the expected ones? If not we should validate them against a well known roles list before sending the request