Skip to content

notebook-example.md

Latest commit

 

History

History
90 lines (68 loc) · 2.41 KB

notebook-example.md

File metadata and controls

90 lines (68 loc) · 2.41 KB

📓 Jupyter Notebook Experience

This file shows you how to use AssemblyAI with Jupyter notebooks, Google Colab, or VS Code notebooks.

🚀 Quick Setup for Notebooks

1. Install Dependencies

!npm install assemblyai dotenv

2. Set API Key

import os
os.environ['ASSEMBLYAI_API_KEY'] = 'your_api_key_here'

3. Import and Initialize

from assemblyai import AssemblyAI

client = AssemblyAI({
    'apiKey': os.environ['ASSEMBLYAI_API_KEY']
})

4. Run Transcription

transcript = await client.transcripts.transcribe({
    'audio': 'https://storage.googleapis.com/aai-web-samples/5_audio_demos/02_meeting_audio_16k.wav',
    'speaker_labels': True,
    'auto_highlights': True,
    'sentiment_analysis': True,
    'auto_chapters': True,
})

📱 Google Colab Ready

Copy and paste this into a Google Colab cell:

# Install AssemblyAI SDK
!npm install assemblyai dotenv

# Set your API key
import os
os.environ['ASSEMBLYAI_API_KEY'] = 'your_api_key_here'

# Import and use
from assemblyai import AssemblyAI
client = AssemblyAI({'apiKey': os.environ['ASSEMBLYAI_API_KEY']})

# Transcribe audio
transcript = await client.transcripts.transcribe({
    'audio': 'https://storage.googleapis.com/aai-web-samples/5_audio_demos/02_meeting_audio_16k.wav',
    'speaker_labels': True,
})

print(transcript.text)

🎯 Sample Audio Files

Use these sample URLs for testing:

  • Meeting: https://storage.googleapis.com/aai-web-samples/5_audio_demos/02_meeting_audio_16k.wav
  • Interview: https://storage.googleapis.com/aai-web-samples/5_audio_demos/01_interview_audio_16k.wav
  • Presentation: https://storage.googleapis.com/aai-web-samples/5_audio_demos/03_presentation_audio_16k.wav

📚 Full Notebook Examples

For complete notebook examples, see:

  • basic-transcription.ipynb - Full transcription workflow
  • speaker-diarization.ipynb - Speaker identification
  • lemur-analysis.ipynb - AI-powered insights

💡 Tips for Notebooks

  1. Use await for async operations
  2. Check status before accessing results
  3. Handle errors gracefully
  4. Use progress indicators for long operations
  5. Save results to variables for reuse

🔧 Troubleshooting

  • Import errors: Make sure you've run the install cell
  • API key issues: Check your environment variable
  • Network errors: Verify internet connection
  • Rate limits: Wait between requests