Upload README.md with huggingface_hub

README.md

@@ -1,12 +1,54 @@
----
-title: Hindi High-Quality TTS
-emoji: 🗣️
-colorFrom: red
-colorTo: green
-sdk: gradio
-sdk_version: 6.10.0
-app_file: app.py
-pinned: false
----
-
-High-quality Hindi Text-to-Speech (TTS) using the **Kokoro-82M** model. Optimized for CPU performance on Hugging Face Spaces.
+---
+title: Hindi Text to Speech - हिंदी TTS
+emoji: 🎙️
+colorFrom: yellow
+colorTo: purple
+sdk: gradio
+sdk_version: 5.23.0
+app_file: app.py
+pinned: true
+license: mit
+short_description: High Quality Hindi Text-to-Speech with 2 AI Models
+---
+
+# 🎙️ Hindi Text-to-Speech (हिंदी टेक्स्ट टू स्पीच)
+
+High-quality Hindi Text-to-Speech application powered by **two AI models**:
+
+## Models
+
+### ⚡ Model 1: Facebook MMS-TTS Hindi
+- **Architecture:** VITS (Variational Inference with adversarial learning for end-to-end Text-to-Speech)
+- **Developer:** Meta / Facebook Research
+- **Size:** ~300MB (lightweight)
+- **Speed:** Fast inference on CPU
+- **Quality:** High-quality neural TTS
+
+### 🎵 Model 2: Silero TTS Indic
+- **Developer:** Silero Team
+- **Size:** ~50MB (ultra-lightweight)
+- **Speed:** Ultra-fast inference on CPU
+- **Features:** Multiple Hindi voices (male & female), adjustable sample rates
+- **Quality:** Good quality, very natural sounding
+
+## Features
+- ✅ No API keys required
+- ✅ Runs on free CPU Space
+- ✅ Two different TTS models for comparison
+- ✅ Hindi Devanagari script input
+- ✅ Multiple voice options
+- ✅ Adjustable audio quality
+- ✅ Beautiful, modern UI
+
+## Usage
+1. Select a model tab (MMS-TTS or Silero TTS)
+2. Type or paste Hindi text in Devanagari script
+3. Adjust voice settings (seed/speaker/sample rate)
+4. Click "Generate Speech"
+5. Listen to and download the generated audio
+
+## Tech Stack
+- Gradio for the web interface
+- HuggingFace Transformers for MMS-TTS
+- PyTorch Hub for Silero TTS
+- Aksharamukha for Hindi transliteration