GhulamMustafa0/children-story-api

📚 Children's Story Generator API

A FREE API for generating personalized children's stories using a fine-tuned Llama 2 model!

🌟 Features

• —🎭 Personalized stories based on child's details
• —🆓 Completely free to use
• —🚀 Simple REST API
• —🎨 Beautiful web interface
• —🔒 Privacy-focused (no data stored)

🚀 Quick Start

Web Interface

Visit the Space and use the interactive interface to generate stories instantly!

API Usage

Python

import requests

API_URL = "https://ghulamustafa0-children-story-api.hf.space/api/predict"

def generate_story(name, age, mood, animal, max_length=500, temperature=0.7):
    response = requests.post(API_URL, json={
        "data": [name, age, mood, animal, max_length, temperature]
    })
    return response.json()["data"][0]

# Generate a story
story = generate_story("Emma", "6", "curious", "fox")
print(story)

JavaScript/Node.js

const API_URL = "https://ghulamustafa0-children-story-api.hf.space/api/predict";

async function generateStory(name, age, mood, animal) {
    const response = await fetch(API_URL, {
        method: "POST",
        headers: { "Content-Type": "application/json" },
        body: JSON.stringify({
            data: [name, age, mood, animal, 500, 0.7]
        })
    });
    
    const result = await response.json();
    return result.data[0];
}

// Usage
generateStory("Emma", "6", "curious", "fox")
    .then(story => console.log(story));

cURL

curl -X POST https://ghulamustafa0-children-story-api.hf.space/api/predict   -H "Content-Type: application/json"   -d '{"data": ["Emma", "6", "curious", "fox", 500, 0.7]}'

📋 API Parameters

📊 Response Format

{
  "data": [
    "Once upon a time, there was a curious girl named Emma who loved foxes..."
  ],
  "duration": 23.5
}

⏱️ Performance

• —First Request: 30-60 seconds (cold start)
• —Subsequent Requests: 15-30 seconds
• —Concurrent Users: Supported via queue system
• —Rate Limit: No hard limits (fair use policy)

🛠️ Technical Details

• —Base Model: Llama 2 7B Chat
• —Fine-tuning: LoRA (Low-Rank Adaptation)
• —Quantization: 4-bit (bitsandbytes)
• —Framework: Transformers, PEFT, Gradio
• —Hardware: CPU (Hugging Face free tier)

💡 Usage Tips

1. 1.First request is slow: The model needs to load (30-60s). Subsequent requests are much faster.
2. 2.Optimal temperature: 0.7 provides good balance between creativity and coherence
3. 3.Story length: 400-600 tokens works best for children's stories
4. 4.Mood options: stick to: happy, curious, brave, excited, scared
5. 5.Animals: Any animal works, but common ones (fox, lion, bear) produce better results

🎯 Use Cases

• —📱 Mobile apps for children
• —🌐 Educational websites
• —📖 E-book generation
• —🎮 Interactive storytelling games
• —👨‍👩‍👧 Parenting apps

🔒 Privacy & Safety

• —No data is stored or logged
• —Stories are generated fresh each time
• —Content is appropriate for children
• —No personal information required

📜 License

This project uses the Llama 2 Community License from the base model.

🤝 Contributing

Feel free to:

• —Report issues
• —Suggest improvements
• —Share your use cases

📧 Support

For questions or issues, please open a discussion on this Space.

🎉 Credits

• —Model: Based on Meta's Llama 2 7B Chat
• —Fine-tuning: Custom children's story dataset
• —Framework: Hugging Face Transformers & Gradio
• —Creator: GhulamMustafa0

Enjoy creating magical stories for children! ✨📚