Add Configurable Timeouts for Operations and Ollama Integration Documentation

[shreyashankar]

This PR introduces configurable timeouts for operations and adds comprehensive documentation for Ollama integration, enhancing our system's reliability and user guidance. This will close #23.

Changes

1. Added support for configurable timeouts in operations
2. Implemented a new max_retries_per_timeout parameter (default: 2)
3. Created a new test in test_map.py to verify the timeout functionality
4. Added documentation for Ollama integration with timeout examples

Details

Timeout Implementation

• Operations now accept a timeout parameter in their configuration
• If a timeout occurs, the operation will retry up to max_retries_per_timeout times

New Parameter

• max_retries_per_timeout: Controls the number of retry attempts after a timeout (default: 2)

Testing

• Added a new test case in test_map.py to ensure proper functionality of the timeout feature

Documentation

• Created new documentation for Ollama integration
• Included example of using timeouts with Ollama operations
• Provided guidance on estimating appropriate timeout values

Usage Example

operations:
  - name: classify_medical_record
    type: map
    timeout: 20000  # Example timeout for Ollama operation
    max_retries_per_timeout: 3
    # ... other configuration ...

Benefits

• Improved handling of long-running operations, especially for Ollama integrations
• Enhanced system stability by preventing indefinite hangs
• Greater control over resource allocation and task management
• Clear guidance for users on integrating Ollama and managing timeouts