nrslib 0.0.1

Standardised Library for the Benchmarking of News Recommenders Systems

nrslib

This library is a modular, end-to-end pipeline for news recommender systems.

Quick Start

Main Technologies

Familarity with the follow tools are recommended to use this library.

PyTorch Lightning - a lightweight PyTorch wrapper for high-performance AI research. Think of it as a framework for organizing your PyTorch code.

Hydra - a framework for elegantly configuring complex applications. The key feature is the ability to dynamically create a hierarchical configuration by composition and override it through config files and the command line.

TorchMetrics - a collection of 80+ PyTorch metrics implementations and an easy-to-use API to create custom metrics.

Lightning-Hydra-Template - A clean and scalable template to kickstart your deep learning project.

Running a model

Import the library

import nrslib

Export default configs

nrslib.export_default_config(“~/configs”)

Choose and edit configs in the open console dialog and start training afterwards

nrslib.start_train(“~/configs”, [“experiment=naml.yaml”])

To test a model, run

nrslib.start_test(“~/configs”, [“experiment=naml.yaml”])

To extend a model or datamodule inherit it

from nrslib.src.models.naml import NAML
class ImprovedNAML(NAML):

Then adjust the configurations to use the new class

_target_: path.to.class.ImprovedNAML