https://github.com/architj6/youtube-llm-summarizer
AI-powered YouTube video summarizer built with Streamlit. This project extracts transcripts from YouTube using YouTubeTranscriptAPI or Whisper, then summarizes them using Groq's LLM. Dockerized for seamless deployment.
https://github.com/architj6/youtube-llm-summarizer
ai-summarization ai-summarizer docker ffmpeg groq-api llm python speech-to-text streamlit video-summarization whisper youtube youtube-api youtube-transcript-api
Last synced: 2 months ago
JSON representation
AI-powered YouTube video summarizer built with Streamlit. This project extracts transcripts from YouTube using YouTubeTranscriptAPI or Whisper, then summarizes them using Groq's LLM. Dockerized for seamless deployment.
- Host: GitHub
- URL: https://github.com/architj6/youtube-llm-summarizer
- Owner: ArchitJ6
- License: mit
- Created: 2025-03-14T14:11:48.000Z (2 months ago)
- Default Branch: main
- Last Pushed: 2025-03-15T14:50:18.000Z (2 months ago)
- Last Synced: 2025-03-15T15:25:31.138Z (2 months ago)
- Topics: ai-summarization, ai-summarizer, docker, ffmpeg, groq-api, llm, python, speech-to-text, streamlit, video-summarization, whisper, youtube, youtube-api, youtube-transcript-api
- Language: Python
- Homepage:
- Size: 5.86 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# 📺 YouTube LLM Summarizer
### 🚀 AI-powered YouTube Video Summarization with Streamlit
This project extracts transcripts from YouTube videos using **YouTubeTranscriptAPI** or **Whisper**, then summarizes them using **Groq's LLM**. The app is built with **Streamlit** and runs in a **Dockerized environment** for easy deployment.
---
## 🚀 Features
✅ Extracts YouTube transcripts automatically
✅ Uses **Whisper** for speech-to-text if transcripts are unavailable
✅ Summarizes content using **Groq LLM**
✅ Provides a simple **Streamlit UI** for easy interaction
✅ **Dockerized** for seamless deployment---
## 📂 Project Structure
```
📁 youtube-llm-summarizer
│── 📄 app.py # Streamlit UI for summarization
│── 📄 summarizer.py # Core logic for transcript extraction & summarization
│── 📄 Dockerfile # Docker configuration
│── 📄 docker-compose.yml # Docker Compose setup
│── 📄 .env # Environment variables
│── 📄 requirements.txt # Python dependencies
│── 📁 temp/ # Temporary storage for audio files
```---
## 🛠️ Installation & Setup
### 1️⃣ **Clone the Repository**
```sh
git clone https://github.com/ArchitJ6/youtube-llm-summarizer.git
cd youtube-llm-summarizer
```### 2️⃣ **Set Up Environment Variables**
Create a `.env` file in the root directory and add your API keys:
```
GROQ_API_KEY=your_groq_api_key
```### 3️⃣ **Run with Docker Compose**
```sh
docker-compose up --build
```
This will:
- Pull the necessary dependencies
- Set up the Streamlit app in a container
- Run the app on **http://localhost:8501**---
## 🎬 How to Use
1. Open **http://localhost:8501**
2. Enter a **YouTube video URL**
3. Click **"Summarize"**
4. View the **generated summary**---
## 📦 Running Without Docker
If you prefer to run locally:
```sh
pip install -r requirements.txt
streamlit run app.py
```---
## 🛠️ Built With
- **Python** 🐍
- **Streamlit** 📊
- **YouTubeTranscriptAPI** 🎥
- **Whisper** 🎙️
- **Groq LLM** 🧠
- **Docker & Docker Compose** 🐳---
## 📝 License
This project is licensed under the [MIT License](LICENSE).