Brusta

• language-agnostic PyTorch model serving
• serve JIT compiled PyTorch model in production environment

Requirements

• docker == 18.09.1
• wget == 1.20.1
• your JIT traced PyTorch model (If you are not familiar with JIT tracing, please refer JIT Tutorial)

Process Flow

1. run bridge

Request Example

request to the model server as follow (suppose your input dimension is 3)

curl -X POST -d '{"input":[1.0, 1.0, 1.0]}' localhost:8080/model/predict

Contributors

• YongRae Jo ([email protected])
• YoonHo Jo ([email protected])
• GiChang Lee ([email protected])
• Seunghwan Hong
• SeungHyek Cho

Author

• Alex Kim ([email protected])