mirror of
https://github.com/open-mmlab/mmdeploy.git
synced 2025-01-14 08:09:43 +08:00
ff39377833
* prepare for split * mark multiclass nms * redesign mark decorator to handle input/output names better (#17) * redesign mark decorator to handle input/output names better * fixed a bug in rename_value; robustify extract_model * fix lint * fix log in deploy pipeline * fix onnxruntime config * fix lint * update connect init * rename is_inspecting * simplify function_mark Co-authored-by: Li Zhang <lzhang329@gmail.com>
MMDeployment
Installation
-
Build backend ops
-
update submodule
git submodule update --init -
Build with onnxruntime support
mkdir build cd build cmake -DBUILD_ONNXRUNTIME_OPS=ON -DONNXRUNTIME_DIR=${PATH_TO_ONNXRUNTIME} .. make -j10 -
Build with tensorrt support
mkdir build cd build cmake -DBUILD_TENSORRT_OPS=ON -DTENSORRT_DIR=${PATH_TO_TENSORRT} .. make -j10 -
Or you can add multiple flags to build multiple backend ops.
-
-
Setup project
python setup.py develop
Usage
python ./tools/deploy.py \
${DEPLOY_CFG_PATH} \
${MODEL_CFG_PATH} \
${MODEL_CHECKPOINT_PATH} \
${INPUT_IMG} \
--work-dir ${WORK_DIR} \
--device ${DEVICE} \
--log-level INFO