Skip to content

Latest commit

 

History

History
35 lines (25 loc) · 1.11 KB

README.md

File metadata and controls

35 lines (25 loc) · 1.11 KB
Raw

Prepare models for TorchServe

Virtual environment

To install and activate venv run

poetry install --with=dev
poetry shell

Yolo

torch-model-archiver --model-name yolo --version 1.0 --serialized-file ../nn/nn/weights/yolo.pt --handler inference_torchserve/handlers/yolo.py --export-path inference_torchserve/model_store/

STN

python inference_torchserve/scripts/prepare_stn.py
torch-model-archiver --model-name stn --version 1.0 --serialized-file /tmp/stn.pt --handler inference_torchserve/handlers/stn.py --export-path inference_torchserve/model_store/

LPRNET

python inference_torchserve/scripts/prepare_lprnet.py
torch-model-archiver --model-name lprnet --version 1.0 --serialized-file /tmp/lprnet.pt --handler inference_torchserve/handlers/lprnet.py --export-path inference_torchserve/model_store/

Plate recognition

torch-workflow-archiver --workflow-name plate_recognition --spec-file inference_torchserve/workflow/plate_recognition.yaml --handler inference_torchserve/handlers/plate_recognition.py --export-path inference_torchserve/model_store/