fix: Resolve issues #7, #8, #9 - moviepy, transformers, Whisper OOM

feat: Add streaming Ollama support, model caching, and UI improvements

fix: Update moviepy import for v2.x compatibility (closes #6)

fix: Free up disk space in GitHub Actions runner

fix: Resolve protobuf dependency conflict for pyannote.audio 4.x

fix: Remove torch from requirements.txt and use flexible versions

fix: Use flexible PyTorch versions and fix Docker build order

fix: Update PyTorch/torchaudio to 2.1.0 for SpeechBrain 1.0 diarization compatibility

Refactor keyword extraction logic to convert stopwords to a list for compatibility with scikit-learn; add checks for empty keywords list in timestamp extraction function.

Enhance file selection to support additional audio format (M4A) and update README.md to reflect new supported formats for video and audio recordings.

Refactor Dockerfile and requirements.txt for improved dependency management; update README.md with clearer installation options and instructions for prebuilt images.

Enhance README.md with Docker installation instructions and update Ollama API endpoint to be configurable via environment variable.

Enhance file selection to support multiple video formats (MP4, AVI, MOV, MKV) and provide user guidance on supported formats.

Fix summarization issues and improve GPU handling. Update .gitignore for venv

Update README.md