marcin-laskowski / ml-fastapi-docker Goto Github PK

The repository for the artcile Deploy ML model using FastAPI, Docker, and Cloud Run where you will learn a step-by-step tutorial on how to deploy a PyTorch implementation of the AlexNet Image classifier.

.
├── app.py                      > main fastapi app
├── docker-compose.yaml
├── Dockerfile
├── README.md
├── requirements.txt            > dependencies
├── sample_data
│   ├── input.jpg               > sample input image
│   └── sample_input.json       > JSON with sample input url
└── syndicai.py                 > file needed to deploy via Syndicai Platform

• Basic understanding of Python, Docker, REST
• Access to Google Cloud Platform (if you want to host a model)
• Trained model from the repository.
• Installed docker, docker-compose, gcloud, python 3.6
• Installed all python dependencies listed below. torch==1.7.1, torchvision==0.8.2, fastapi==0.63.0, uvicorn==0.13.3, requests==2.18.4

If you found that material helpful, or have some questions, please let us know via mail or slack.

built with passion by Syndicai.