Update Space configuration and documentation

.gitignore

@@ -0,0 +1,53 @@
+# Python
+__pycache__/
+*.py[cod]
+*$py.class
+*.so
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+
+# Virtual Environment
+venv/
+ENV/
+env/
+
+# IDE
+.idea/
+.vscode/
+*.swp
+*.swo
+
+# Logs
+*.log
+
+# Local development
+.env
+.env.local
+
+# Model weights
+weights/
+
+# Generated content
+outputs/
+temp/
+*.mp4
+*.wav
+
+# Keep example files
+!examples/*.json
+!assets/examples/*
+!assets/audio/* 

README.md

@@ -1,61 +1,74 @@
 ---
-title: MeiGen MultiTalk Demo
+title: Phunter Space - Video Generation Demo
 emoji: 🎬
 colorFrom: blue
-colorTo: red
+colorTo: purple
 sdk: gradio
-sdk_version: 4.19.2
+sdk_version: 4.12.0
 app_file: app.py
 pinned: false
 license: apache-2.0
-hf_oauth: true
-models:
-  - MeiGen-AI/MeiGen-MultiTalk
-  - TencentGameMate/chinese-wav2vec2-base
-tags:
-  - audio
-  - video
-  - image
-  - text-to-video
 ---
 
-# MeiGen-MultiTalk
+# Phunter Space - Video Generation Demo
 
-Audio-driven multi-person conversational video generation system based on [MeiGen-AI/MeiGen-MultiTalk](https://huggingface.co/MeiGen-AI/MeiGen-MultiTalk).
+This is a Gradio demo for generating talking head videos from images and audio using advanced AI models.
 
-## Features
+## 🌟 Features
 
-- 💬 Realistic Conversations - Support single & multi-person generation
-- 👥 Interactive Character Control - Direct virtual humans via prompts
-- 🎤 Generalization Performance - Support generation of cartoon characters and singing
-- 📺 Resolution Flexibility - 480p & 720p output at arbitrary aspect ratios
-- ⏱️ Long Video Generation - Support videos up to 15 seconds
+- 💬 Generate talking head videos from images and audio
+- 👥 Support for both single and multi-person video generation
+- 🎯 High-quality lip synchronization
+- 📺 Support for multiple resolutions (480p, 720p)
+- 🎨 Customizable generation parameters
 
-## Setup
+## 🚀 Quick Start
 
-1. Install dependencies:
-```bash
-pip install -r requirements.txt
-```
+1. Click "Load Night Studio Example" or "Load Day Studio Example"
+2. Upload your audio file (WAV format)
+3. Click "Generate Video"
 
-2. Download required models:
-```bash
-huggingface-cli download MeiGen-AI/MeiGen-MultiTalk --local-dir ./weights/MeiGen-MultiTalk
-huggingface-cli download TencentGameMate/chinese-wav2vec2-base --local-dir ./weights/chinese-wav2vec2-base
-```
+## 📝 Parameters Guide
 
-## Usage
+### Resolution
+- 480p: Faster generation, lower quality
+- 720p: Better quality, slower generation
 
-See the examples directory for sample configurations:
-- `examples/single_example.json` - Single person video generation
-- `examples/multi_example.json` - Multi-person conversation generation
+### Audio CFG (1.0-10.0)
+- Controls lip movement influence
+- Recommended: 4.0
+- Higher values = more pronounced articulation
 
-## License
+### CFG Scale (1.0-15.0)
+- Controls prompt adherence
+- Recommended: 7.5
+- Higher values = stricter prompt following
 
-This project is licensed under the Apache License 2.0 - see the LICENSE file for details.
+### Max Duration
+- Limits output video length
+- Maximum: 15 seconds
+- Default: 10 seconds
 
-## Configuration Options
+## 💡 Tips
 
-- `image`: Path to reference image
-- `audio`: Path to audio file(s)
-- `
+1. Use high-quality reference images
+2. Provide detailed prompts
+3. Start with example settings
+4. Experiment with CFG values
+5. Ensure good lighting in reference images
+
+## 📋 Requirements
+
+- Input Image: Clear face photo(s)
+- Audio: WAV format
+- Prompt: Detailed scene description
+
+## 🛠 Technical Details
+
+- Model: MeiGen MultiTalk
+- Framework: Gradio 4.12.0
+- GPU: T4 (recommended)
+
+## 📬 Contact
+
+For questions or issues, please visit the [GitHub repository](https://github.com/yourusername/phunter_space) or create an issue on Hugging Face Spaces.