README.md

---
title: YouTube Translator & Speaker
emoji: 🎤
colorFrom: indigo
colorTo: blue
sdk: docker
app_port: 7860
pinned: false
---

# YouTube Translator & Speaker (Hugging Face — Docker Space)

A fully automated YouTube translation and speech tool that:
✅ Extracts transcripts using YouTube captions  
✅ Uses OpenAI Whisper if captions are missing (requires `OPENAI_API_KEY`)  
✅ Translates into a selected language using OpenAI GPT  
✅ Can optionally generate speech using `edge-tts` (no key required)  

---

## 🚀 Deploy on Hugging Face (Docker)

1) Create a new Space → **Docker**  
2) Upload these files: `Dockerfile`, `app.py`, `requirements.txt`, `README.md`  
3) Go to **Settings → Repository secrets → Add new secret**  
   - Name: `OPENAI_API_KEY`  
   - Value: `sk-xxxxxxxxxxxxxxxxxxxxx`  
4) Click **Restart/Build Space**

---

## ▶️ Use the App
1. Paste a **YouTube URL** or **11-char Video ID**  
2. Select target language (`en`, `es`, `hi`, `ha`, etc.)  
3. (Optional) Keep **Generate Speech** enabled  
4. Click **Submit** → read translated text / play MP3

---

## ⚠️ Common Issues

| Issue | Cause | Solution |
|------|------|---------|
| Invalid video | Bad URL/ID | Use full URL or exact 11-char ID |
| No captions & no key | Video lacks subtitles | Add `OPENAI_API_KEY` for Whisper fallback |
| Translation failed | Key missing/expired | Re-add correct key in Secrets |
| TTS failed | Network hiccup | Retry or uncheck speech |

---

## 💻 Local Run (optional)

```bash
python -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt
export OPENAI_API_KEY=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
python app.py