mirrors
/
mmdeploy
mirror of https://github.com/open-mmlab/mmdeploy.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity
mmdeploy/demo
History
Chen Xin 0ce7c83c63
mmrotate sdk module (#450) 
* support mmrotate

* fix name

* windows default link to cudart_static.lib, which is not compatible with static build && python_api

* python api

* fix ci

* fix type & remove unused meta info

* fix doxygen, add [out] to @param

* fix mmrotate-c-api

* refactor naming

* refactor naming

* fix lint

* fix lint

* move replace_RResize -> get_preprocess

* Update cuda.cmake

On windows, make static lib and python api build success.

* fix ptr

* Use unique ptr to prevent memory leaks

* move unique_ptr

* remove deleter

Co-authored-by: chenxin2 <chenxin2@sensetime.com>
Co-authored-by: cx <cx@ubuntu20.04>
 		2022-05-17 23:37:32 +08:00
..
csrc mmrotate sdk module (#450) 2022-05-17 23:37:32 +08:00
resources [Feature]: Support mmpose (#94) 2022-02-16 11:03:12 +08:00
README.md add a demo for rewrite (#145) 2021-10-27 11:06:42 +08:00
demo_rewrite.py Add a new pre-commit-hook to automatically add a copyright (#166) 2022-02-23 11:38:40 +08:00

README.md

Demo

We provide a demo showing what our mmdeploy can do for general model deployment.

In demo_rewrite.py, a resnet18 model from torchvision is rewritten through mmdeploy tool. In our rewritten model, the forward function of resnet gets modified to only down sample the original input to 4x. Original onnx model of resnet18 and its rewritten are visualized through netron.

Prerequisite

Before we run demp_rewrite.py, we need to install pyppeteer through:

pip install pyppeteer

Demo results

The original resnet18 model and its modified one are visualized as follows. The left model is the original resnet18 while the right model is exported after rewritten.

Original resnet18 Rewritten model