v0.1.7

[0.1.7] 2025-06-08

Added

• OpenAI API support: Added OpenAI as an alternative summarization provider alongside existing Gemini support
• Provider selection option: New --summarize-by command-line argument with choices ["openai", "gemini"]
• Web UI provider selector: Added dropdown in Streamlit interface to choose between OpenAI and Gemini
• OpenAI utility functions:
  • get_openai_prompt_parts() for message formatting
  • get_openai_default_config() for API configuration
• Comprehensive test suite: Added unit tests for both OpenAI and Gemini functionality with proper mocking
• Enhanced error handling: Provider-specific API key validation and error messages

Fixed

• Async compatibility: Resolved async/sync compatibility issues in summarization workflow

Changed

• Default summarization provider: Changed from Gemini to OpenAI as the default option
• Async summarization: Updated _summarize() function to be fully async for better performance
• Environment variable handling: Enhanced API key validation logic for multiple providers
• Documentation: Updated README with new command-line options and usage examples
• Project structure: Added file_dump/ to .gitignore for better repository hygiene

v0.1.6

[0.1.6] 2025-05-01

Added

• Docker support
• Automatic file cleanup functionality (purger)
• Added startup scripts
• Added schedule dependency

Changed

• Removed unnecessary dependencies (transformers, datasets, accelerate, etc.)
• Updated Poetry configuration, added script entry points

Fixed

• Optimized log output in Docker environment

v0.1.5

Added

• Offline APP (work w/ local Whisper)

v0.1.5-rc1

Added

• Add streamlit UI.

v0.1.4

[0.1.4] - 2024-06-14

Added

• Add option duration to control the size of split audio.
• User check if the split file size exceeds 25 MB (limit from Whisper).

Changed

• Use google-generativeai instead of requesting manually.

Fixed

• Remove the tmp files while API request interrupted.

Deprecated

• Request manually via url.

v0.1.4-rc1

[0.1.4-rc1] - 2024-06-03

Added

• Add option duration to control the size of split audio.
• User check if the split file size exceeds 25 MB (limit from Whisper).

Changed

• Use google-generativeai instead of requesting manually.

Fixed

• Remove the tmp files while API request interrupted.

Deprecated

• Request manually via url.

v0.1.3

Add:

• Improve log message

v0.1.2

Fixes:

• spaces between file name.

Add:

• --lang to assign responded language.
• Request Whisper with AsyncOpenAI

v0.1.1

Fixes:

• Remove redundant dumping of data.json.

Add:

• Add mode for summary only.
• Add some log messages.