Web API that summarizes multimedia from various sources using modern AI tools.
flowchart LR
API[HTTP API]
Whisper[whisper.cpp]
Llama[llama.cpp]
Output
URL --> API
API -->|Audio File| Whisper
Whisper -->|Transcript| Llama
Llama --> |Summary| Output
- YouTube
- Apple Podcasts
- Direct file URLs
- Python 3.11
- Poetry
- FFmpeg
- whisper.cpp
- llama.cpp
- llama.cpp compatible model
-
Clone the Repository:
git clone https://github.com/oscargullberg/tldwol/blob/main/README.md cd tldwol
-
Install Dependencies:
poetry install
Create a .env
file in the project root and add your environment variables:
LLAMA_CPP_DIR_PATH=path/to/llama.cpp
WHISPER_CPP_DIR_PATH=path/to/whisper.cpp
FILES_DL_DIR_PATH=path/to/download/directory
LLAMA_MODEL_PATH=path/to/model
WHISPER_MODEL_PATH=/some/path/bin/whisper.cpp/models/ggml-large.bin
-
Start the Server:
poetry run start
-
Request a Summary:
curl "http://127.0.0.1:8000/?url=https://youtube.com/watch?v=thOifuHs6eY"
{ "summary": "The provided content highlights the significance of hexagons in nature and human-made objects. It emphasizes the strength and stability of hexagons and ..." }