This is a comprehensive library designed to facilitate various machine learning projects using PyTorch. It provides essential functionalities such as custom layers, dataset handling, and utility functions for model training and visualization.
- Custom layers for pixel normalization and upsampling/downsampling.
- Convenient data transformation and augmentation functions.
- Custom dataset class for handling image datasets.
- Utility functions for model initialization, visualization, and more.
This tool requires Python. Use this command to install the library:
pip install polip
Make sure to install the following required libraries:
pip install matplotlib os torch PIL numpy torchvision
The library includes custom layers like PixelNormLayer
, UpSample
, and DownSample
. Here's an example of how to use them:
from polip.cb import PixelNormLayer, UpSample, DownSample
You can use the CustomImageDataset
class to handle image datasets:
from polip import CustomImageDataset, get_rgb_transform