https://github.com/myatmyintzuthin/libtorch-inference

https://github.com/myatmyintzuthin/libtorch-inference

Last synced: 16 days ago
JSON representation

• Host: GitHub
• URL: https://github.com/myatmyintzuthin/libtorch-inference
• Owner: myatmyintzuthin
• License: apache-2.0
• Created: 2023-02-07T04:37:01.000Z (almost 2 years ago)
• Default Branch: main
• Last Pushed: 2024-10-27T10:15:42.000Z (about 2 months ago)
• Last Synced: 2024-10-27T11:38:18.746Z (about 2 months ago)
• Language: C++
• Size: 202 KB
• Stars: 1
• Watchers: 1
• Forks: 0
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# torchlib-inference

![Ubuntu](https://img.shields.io/badge/Ubuntu-E95420?style=for-the-badge&logo=ubuntu&logoColor=white)

![Torch](https://img.shields.io/badge/PyTorch-EE4C2C?style=for-the-badge&logo=pytorch&logoColor=white)

![C++](https://img.shields.io/badge/C%2B%2B-00599C?style=for-the-badge&logo=c%2B%2B&logoColor=white)

![Cmake](https://img.shields.io/badge/CMake-064F8C?style=for-the-badge&logo=cmake&logoColor=white)

This repo consists of C++ libtorch inference implementation for pytorch classification model.

### Install Dependencies

1. Download libtorch distribution from [official website](https://pytorch.org/get-started/locally/). libtorch version used in this repo [link](https://download.pytorch.org/libtorch/cpu/libtorch-cxx11-abi-shared-with-deps-2.0.0%2Bcpu.zip).

2. Unzip the downloaded file and store `libtorch` folder in `external_packages` folder. If `external_packages` folder doesn't exit, create one.

3. Install OpenCV: follow the installation instructions in this [guide](https://github.com/myatmyintzuthin/extract-table/blob/C%2B%2B/Installation_Guide.md).

### Run inference

- Before running inference, the pytorch model should be converted to script model. Reference, [trace_model.py](https://github.com/myatmyintzuthin/jpfood-classification-pytorch/blob/master/trace_model.py).

- Build cmake:

```

mkdir build

cd build

cmake -DCMAKE_PREFIX_PATH=/path/to/libtorch ..

cmake --build . --config Release

```

- run inference, inside `build` folder: 

```

./libtorch --model_path ../script_model/resnet_food_script.pt --image_path ../test-images/sushi.jpg

```

### Results

```

Finish model loading ... 

Finish image processing ... 

---------------------------------------

Predicted class = Sushi | confidence = 0.999214

```