Add Automated OmniParser Deployment

[abrichr]

Adds automated deployment for OmniParser to EC2:

• Dockerfile with CUDA support
• Deployment automation script
• Environment configuration
• Setup documentation
• Client

Tested:

• Docker build
• Model loading
• API endpoints

(deploy) abrichr@MacBook-Pro-5 deploy % python deploy/models/omniparser/deploy.py start
...
2025-02-18 15:40:12.670 | INFO     | __main__:execute_command:289 - Executing: curl -s http://localhost:8000/probe/
2025-02-18 15:40:12.915 | ERROR    | __main__:execute_command:333 - Command failed with exit status 56: curl -s http://localhost:8000/probe/
2025-02-18 15:40:12.915 | WARNING  | __main__:start:431 - Server not ready (attempt 11/30): Command failed with exit status 56: curl -s http://localhost:8000/probe/
2025-02-18 15:40:12.915 | INFO     | __main__:start:433 - Waiting 10 seconds before next attempt...
2025-02-18 15:40:22.920 | INFO     | __main__:execute_command:289 - Executing: curl -s http://localhost:8000/probe/
2025-02-18 15:40:23.030 | INFO     | __main__:execute_command:302 - {"message":"Omniparser API ready"}
2025-02-18 15:40:23.036 | INFO     | __main__:execute_command:336 - Successfully executed: curl -s http://localhost:8000/probe/
2025-02-18 15:40:23.036 | INFO     | __main__:execute_command:289 - Executing: docker ps | grep omniparser-container
2025-02-18 15:40:23.201 | INFO     | __main__:execute_command:302 - 7499776f1a14   omniparser   "/opt/nvidia/nvidia_…"   2 minutes ago   Up 2 minutes   0.0.0.0:8000->8000/tcp, :::8000->8000/tcp   omniparser-container
2025-02-18 15:40:23.214 | INFO     | __main__:execute_command:336 - Successfully executed: docker ps | grep omniparser-container
2025-02-18 15:40:23.214 | INFO     | __main__:start:443 - Deployment complete. Server running at: http://98.84.105.209:8000
2025-02-18 15:40:23.216 | WARNING  | __main__:start:454 - Could not verify external access: No module named 'requests'
2025-02-18 15:40:23.216 | INFO     | __main__:start:477 - Deployment completed successfully!

(deploy) abrichr@MacBook-Pro-5 deploy % python deploy/models/omniparser/client.py --image_path ../tests/assets/excel.png --server_url http://3.238.224.89:8000
Server is available
API Latency: 3.00 seconds