Pytorch Knowledge

Pytorch Learning Manual

Most machine learning workflows involve processing data, creating models, optimizing model parameters, and saving the trained model. This tutorial introduces you to a complete ML workflow implemented in PyTorch and provides links to learn more about each concept.

This tutorial will cover the basic knowledge of machine learning and deep learning, covering回归分类问题and many well-known basic frameworks of deep learning, such as CNN , RNN , fastai , etc.

The official Facebook tutorial catalog will use the FashionMNIST dataset to train a neural network that predicts whether an input image belongs to one of the following categories: T 恤/上衣,裤子,套头衫,连衣裙,外套,凉鞋,衬衫,运动鞋,包, or脚踝.

The Pytorch deep learning practical catalog will start with the simplest linear regression problems and gradually transition to complex deep learning models.

This tutorial assumes you have basic familiarity with Python and deep learning concepts. The document (Facebook official tutorial) is translated from Pytorch official website

Run the tutorial code

You can run this tutorial in several ways:

• in the cloud
• Local: This option requires you to first set up PyTorch and TorchVision on your local machine (installation instructions). Copy the code into your favorite IDE.

Facebook official tutorial directory

1. Tensors
2. Datasets and DataLoaders
3. Transforms

Pytorch deep learning practical directory

1. Linear and nonlinear regression problems
2. logistic regression problem
3. CNN in action: Handwritten font recognition
4. Transfer learning: Using AlexNet to implement cat and dog classification
5. RNN practice: using LSTM to achieve function prediction
6. GAN practice: Use GAN to generate Shiba Inu avatars