Fix issue #2044: Improve output parsing and timeout for local LLMs

[jjmachan]

This PR addresses the issues reported in #2044 where local LLMs like Ollama with qwen2.5:14b were encountering RagasOutputParserException and TimeoutError during evaluation.

Changes

1. Increased default timeout in RunConfig from 180 to 300 seconds to accommodate slower local LLMs
2. Enhanced extract_json function to better handle various LLM output formats:
   • Added support for markdown code blocks
   • Added handling for single quotes in JSON
   • Added handling for trailing commas in JSON
   • Improved JSON extraction from text with surrounding content
3. Improved RagasOutputParser with:
   • More robust error handling
   • Increased default retries from 1 to 3
   • Better error messages with details about the failure
   • Added suggestions for working with local LLMs in error messages

Added tests

• Tests for extract_json with various input formats
• Tests for RagasOutputParser fallback mechanism
• Tests for timeout configuration

Fixes #2044

@jjmachan can click here to continue refining the PR