jjerry-k / triton_sample

Triton Inference Server Example

Quick Setting

Quick Setting run on cpu.
Model format is TorchScript.

1. Download Model

• Download Link: https://drive.google.com/drive/folders/1QeKk-7v-JzyG0v8wquybT7gjJatsg4EJ?usp=sharing

2. Move model.pt file to model_repository/{model name}/1

3. Run docker compose

docker compose up

To Do List

• GPU Mode
• Detection router
  • Postprocessing
• Segmentation router
• Variable input type
• Variable output type