gradio-demo

Gradio Demo on Akash

Deploy interactive machine learning demos and data apps using Gradio on Akash Network's decentralized cloud.

What is Gradio?

Gradio is a Python library that allows you to quickly create customizable web interfaces for machine learning models, APIs, or any Python function. Perfect for:

• Demoing ML models
• Creating interactive data visualizations
• Building simple web tools without learning web development
• Prototyping AI applications

What's Included

This template deploys a simple Gradio application with two demo interfaces:

1. Text Analysis - Mock sentiment analysis demo
2. Image Captioning - Mock image description generator

You can easily customize app.py to add your own models or functions.

Deployment

Prerequisites

• Akash CLI installed
• Akash wallet with ACT tokens
• Basic familiarity with Akash deployments

Deploy on Akash

1. Clone this repository:

   git clone https://github.com/akash-network/awesome-akash.git
   cd awesome-akash/gradio-demo

2. Deploy using Akash CLI:

   akash tx deployment create deploy.yaml --from <your-wallet> --node https://rpc.akashnet.net:443 --chain-id akashnet-2

3. Get your deployment URL: After deployment, check your lease status to find the public URL where your Gradio app is accessible.

Deploy on Akash Console

1. Go to console.akash.network
2. Click "Deploy" → "Build Your Template"
3. Upload the deploy.yaml file
4. Configure resources and pricing
5. Sign and deploy
6. Access your app via the provided URL

Customization

Add Your Own Model

Edit app.py to replace the demo functions with your actual ML models:

import gradio as gr
from transformers import pipeline

# Load your model
classifier = pipeline("sentiment-analysis")

def analyze_sentiment(text):
    result = classifier(text)[0]
    return f"Sentiment: {result['label']}\nConfidence: {result['score']:.2%}"

demo = gr.Interface(
    fn=analyze_sentiment,
    inputs=gr.Textbox(label="Enter text"),
    outputs=gr.Textbox(label="Result"),
    title="Sentiment Analysis on Akash"
)

demo.launch(server_name="0.0.0.0", server_port=7860)

Adjust Resources

Modify the deploy.yaml file to allocate more CPU/RAM if you're running larger models:

resources:
  cpu:
    units: 2.0  # Increase for heavier workloads
  memory:
    size: 4Gi   # Increase for larger models

Cost

Estimated cost on Akash: ~$5-15/month depending on provider and resource allocation.

Significantly cheaper than AWS, Google Cloud, or Heroku for similar workloads.

Resources

• Gradio Documentation
• Akash Network Documentation
• Akash Console

Example Use Cases

• Deploy Hugging Face models for text generation, translation, or classification
• Create data visualization dashboards
• Build interactive demos for research papers or projects
• Host ML model APIs accessible from anywhere

Support

For issues or questions:

• Akash Discord: discord.akash.network
• Gradio Community: huggingface.co/spaces