# Deep-learning-project-Combat-1

**Repository Path**: cdkeeper/deep-learning-project-combat-1

## Basic Information

- **Project Name**: Deep-learning-project-Combat-1
- **Description**: 这是一个简单的深度学习实战项目，你甚至可以零基础去使用它  
- **Primary Language**: Python
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2021-11-01
- **Last Updated**: 2022-02-15

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Deep-learning-project-Combat-1
# # # #
It's a simple deep learning field project, and you can even use it from scratch,
Recently, I have been studying courses related to deep learning, and then I made such a small project. I feel it is easier to understand than many projects, so I can share it with you
I feel very annoyed that I have not been able to get started before. I hope I can give some help to those who have the same experience.
Code environment: PyTorch
File format: First let's create a new folder named work1 as the project folder
Then create four new folders named dataset IMGS logs_train model_weight
Put the core code in the same directory as these folders.
Dataset: used to store dataset. Here, CIFAR-10 is used as an example. Because it is small and convenient to use, there is no need to download dataset.
Model_weight: When we train a network, the most important thing is of course to keep the parameters, put them in this folder, corresponding to the results of each round, save in.pth format
Imgs: contains test images for testing network effects.
Logs_train: is a events file saved by tensorboard, which is convenient for viewing each result on tensorboard. You can create a folder and not use it
Train_gpu_1.py is the same as train_gpu_2.py, but written differently.
Use method 1: run train.py, train_gpu_1, train_gpu_2, and then check the result by using test.py.
Use method 2: If you just want to see the results first, place the mymodule_29.pth file in the model_weight folder, call test.py directly to see the results, and then try training using methods like train
The sample version has been uploaded as a.zip file, so you can see the general structure.
#### Software Architecture
Software Architecture Description
#### Installation Tutorial
1. The pycharm
2. Pytorch
#### Instructions
Use method 1: run train.py, train_gpu_1, train_gpu_2, and then check the result by using test.py.
Use method 2: If you just want to see the results first, place the mymodule_29.pth file in the model_weight folder, call test.py directly to see the results, and then try training using methods like train