# JNU_lab_experiment4

**Repository Path**: arantiska/JNU_lab_experiment4

## Basic Information

- **Project Name**: JNU_lab_experiment4
- **Description**: 江南大学智能计算系统第四次实验：模型推理优化 配套代码
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: release
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 4
- **Created**: 2025-05-15
- **Last Updated**: 2025-05-29

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# PyTorch Custom Convolution Layer

This project is about how to define a custom convolution layer in PyTorch, and use CUDA function to implement convolution.

## Content

[/cpp](https://github.com/Qwesh157/pytorch_custom_convolution_layer/blob/main/cpp) C++ extension include CUDA interface and Python module bind.

[/cuda](https://github.com/Qwesh157/pytorch_custom_convolution_layer/tree/main/cuda) Implicit gemm convolution implementation.

[/include](https://github.com/Qwesh157/pytorch_custom_convolution_layer/tree/main/include) Declaration about forward/backward convolution.

[/pytorch](https://github.com/Qwesh157/pytorch_custom_convolution_layer/tree/main/pytorch) Include setup.py script, custom convolution layer definition.

## Build

```bash
$ sh setup.sh
```

If you don't have root permission, add environment option `--prefix="/home/user/.conda/envs/yourenvname/`.

## TEST

```bash
$ cd pytorch
vim test.py
```


## RUN

run inference program

```
$ python inference.py
```