Hugging Face's logo

Spaces:
shivam98
/
tubeMate
Sleeping

App Files
xet
Community
tubeMate / DEPLOYMENT.md
ShivamPansuriya
Add application file
74708f4
preview code
|
raw
history blame contribute delete
6.63 kB

A newer version of the Gradio SDK is available: 5.49.1

Upgrade

Deployment Guide

This guide covers deploying the Video Transcription Service to Render.com's free tier.

Prerequisites

  1. GitHub Account: Your code needs to be in a GitHub repository
  2. Render Account: Sign up at render.com (free)
  3. Git: Installed on your local machine

Step-by-Step Deployment

1. Prepare Your Repository 

# Initialize git repository (if not already done)
git init

# Add all files
git add .

# Commit changes
git commit -m "Initial commit - Video Transcription Service"

# Add your GitHub repository as remote
git remote add origin https://github.com/yourusername/your-repo-name.git

# Push to GitHub
git push -u origin main

2. Deploy to Render

  1. Go to Render Dashboard

  2. Create New Web Service

    • Click "New +" button
    • Select "Web Service"
    • Choose "Build and deploy from a Git repository"

  3. Connect Repository

    • Select your GitHub repository
    • Click "Connect"

  4. Configure Service

    • Name: video-transcription-service (or your preferred name)
    • Environment: Docker
    • Region: Choose closest to your users
    • Branch: main
    • Dockerfile Path: ./Dockerfile

  5. Advanced Settings

    • Plan: Free (automatically selected)
    • Environment Variables: None needed (auto-configured)
    • Health Check Path: /health
    • Auto-Deploy: Yes (recommended)

  6. Deploy

    • Click "Create Web Service"
    • Render will start building your service

3. Monitor Deployment

  1. Build Process

    • Watch the build logs in real-time
    • First build takes 5-10 minutes (installing dependencies)
    • Look for "Build successful" message

  2. Deployment Status

    • Service will show "Live" when ready
    • Initial startup may take 30-60 seconds (loading AI model)

  3. Test Your Service

    • Your service URL: https://your-service-name.onrender.com
    • API docs: https://your-service-name.onrender.com/docs
    • Health check: https://your-service-name.onrender.com/health

Configuration Details

Automatic Configuration

The service is pre-configured for Render's free tier:

  • Port: Automatically uses $PORT environment variable
  • Memory: Optimized for 512MB limit
  • CPU: Efficient processing for shared CPU
  • Storage: No persistent storage (in-memory only)
  • Health Checks: Configured at /health endpoint

Free Tier Limitations

Resource Limits:

  • 512MB RAM
  • Shared CPU
  • 750 hours/month (service sleeps after 15min inactivity)
  • No persistent storage

Service Behavior:

  • Cold Starts: 30-60 seconds after sleep
  • File Size: 100MB maximum per video
  • Processing: Sequential (one video at a time)
  • Storage: 3.5 hours maximum per transcription

Troubleshooting

Common Build Issues

  1. Out of Memory During Build

    Error: Process killed (out of memory)
    • This is rare but can happen with large dependencies
    • Try pushing smaller commits
    • Contact Render support if persistent

  2. FFmpeg Installation Failed

    E: Unable to locate package ffmpeg
    • Check Dockerfile has correct apt-get commands
    • Ensure base image is correct (python:3.11-slim)

  3. Python Package Installation Failed

    ERROR: Could not install packages
    • Check requirements.txt syntax
    • Ensure all package names are correct
    • Try removing version pins if needed

Runtime Issues

  1. Service Won't Start

    • Check runtime logs in Render dashboard
    • Look for Python import errors
    • Verify all dependencies are installed

  2. Health Check Failing

    Health check failed
    • Service might be taking too long to start
    • Check if Whisper model is loading correctly
    • Verify /health endpoint is accessible

  3. Out of Memory at Runtime

    Process killed (signal 9)
    • Large video files can cause this
    • Reduce MAX_FILE_SIZE in config.py
    • Use smaller Whisper model (tiny instead of base)

  4. Slow Processing

    • First request loads AI model (30-60 seconds)
    • Subsequent requests are faster
    • Consider using smaller model for speed

Service Sleeping

Free Tier Behavior:

  • Service sleeps after 15 minutes of inactivity
  • First request after sleep takes 30-60 seconds
  • This is normal for free tier

Solutions:

  • Upgrade to paid plan for always-on service
  • Use external monitoring to keep service awake
  • Inform users about potential cold start delays

Monitoring and Maintenance

Logs

Access logs in Render dashboard:

  1. Go to your service
  2. Click "Logs" tab
  3. Monitor for errors and performance

Metrics

Monitor service health:

  • Response times
  • Error rates
  • Memory usage
  • Active transcriptions

Updates

Deploy updates automatically:

  1. Push changes to GitHub
  2. Render auto-deploys from main branch
  3. Monitor deployment in dashboard

Scaling Considerations

Free Tier Optimization

Current Setup:

  • Single instance
  • 512MB RAM
  • Shared CPU
  • In-memory storage

Optimization Tips:

  • Use smaller Whisper model for speed
  • Implement request queuing
  • Add request size validation
  • Monitor memory usage

Upgrade Path

Paid Plans Offer:

  • More RAM (1GB+)
  • Dedicated CPU
  • Always-on service
  • Multiple instances
  • Persistent storage options

Security

Current Security Features

  • Rate limiting (10 requests/minute)
  • File size validation
  • File type validation
  • No persistent file storage
  • Automatic cleanup

Additional Security (Optional)

  • API key authentication
  • HTTPS only (automatic on Render)
  • Request logging
  • IP whitelisting
  • CORS configuration

Support

Getting Help

  1. Render Support

    • Free tier includes community support
    • Check Render documentation
    • Use Render community forum

  2. Service Issues

    • Check service logs first
    • Verify configuration
    • Test with smaller files

  3. API Issues

    • Use /docs endpoint for testing
    • Check request format
    • Verify file types and sizes

Useful Commands 

# Test your deployed service
curl https://your-service.onrender.com/health

# Upload test video
curl -X POST "https://your-service.onrender.com/transcribe" \
  -F "[email protected]"

# Check transcription status
curl "https://your-service.onrender.com/transcribe/1"

Your service is now live and ready to transcribe videos! 🎉

Share your service URL with users or integrate it into your applications.