PerfectFrameAI v2.0.0-alpha

This pull request signifies a significant milestone for our project as we prepare to release version v2.0-alpha. It encompasses a broad range of updates aimed at enhancing functionality, improving code documentation, and refining our testing processes. Below is a summary of the key changes included:

Enhancements and New Features:

• Implemented asynchronous image saving to boost performance.
• Added comprehensive testing for various modules including best_frames_extractor, cuda_check, raters, and more.
• Refactored and improved schemas, image evaluators, and docker management.
• Introduced new docstrings across multiple scripts (app, start.py, config.py, image_evaluators.py) to improve code readability and * maintainability.
• Added --build flag logic to docker_manager for enhanced Docker operation.
• Automated service shutdown after task completion to streamline operations.

Bug Fixes:

• Resolved bugs in docker_manager, top_images_extractor, and schemas.py to enhance stability and performance.
• Addressed issues in test scripts post-refactoring to ensure robustness.

Testing:

• Expanded our testing suite to include new e2e tests, integration tests for docker_manager and service_manager, and additional tests across various modules.
• Implemented normalize_images and InceptionResNetNIMA testing to ensure model efficacy and reliability.

Documentation and Project Structure:

• Transitioned from setup.py to start.py, enhancing project setup and configuration.
• Added poetry.lock and pyproject.toml to align with best practices in Python project dependency management.
• Updated README documentation to remove obsolete content and enhance user guidance with new GIF demos.