Tool to help video description generation for multi purposes.
# Clone repository
git clone <repository-url>
cd llm-video-description
# Install the dependencies
sudo apt update && sudo apt install ffmpeg
pip install -r requirements.txt
# ollama serve
curl -fsSL https://ollama.com/install.sh | sh
# Install llama3.1
ollama pull llama3.1
# Run llama3.1
ollama serve
# Run the script
python source/scripts/audio_to_text.py
- 1 - Create the folder llm-video-description/source/audio/, llm-video-description/source/output/ and import your audio file to this new path
- 1.1 - Change the audio file name to transcription.mp3
- 2 - Run the commands in the root folder:
# Run the script
python source/scripts/audio_to_text.py
To:
bind: address already in use
Do:
systemctl stop ollama
ollama serve
- LMM Speech to text: DONE
- LLM to generate description: DONE
- Easy to use: ---