From d2f05dfc5f5eac2622de4d8e9e8b0b9e79149d60 Mon Sep 17 00:00:00 2001 From: maniospas Date: Mon, 19 Aug 2024 11:36:10 +0300 Subject: [PATCH] cleaner readme --- README.md | 46 +++++++++------------------------------------- docs/index.html | 3 ++- 2 files changed, 11 insertions(+), 38 deletions(-) diff --git a/README.md b/README.md index 86cb5fc..976bc63 100644 --- a/README.md +++ b/README.md @@ -1,32 +1,23 @@ # JGNN -A native Java library for graph neural networks. - -**Current requirements: Java 17+** - -## :dart: About +[![Latest nightly release](https://jitpack.io/v/MKLab-ITI/JGNN.svg)](https://jitpack.io/#MKLab-ITI/JGNN) [![Latest stable release](https://img.shields.io/badge/JAR-1.2.0%20stable-red)](https://github.com/MKLab-ITI/JGNN/releases/latest) Graph Neural Networks (GNNs) are getting more and more popular, for example to make predictions based on relational information, or to perform inference -on small datasets. JGNN provides cross-platform implementations of this machine -learning paradigm that do not require dedicated hardware or firmware. +on small datasets. JGNN provides native Java implementations of this machine +learning paradigm, and does not require dedicated hardware or firmware. +Follow the Jitpack badge for Gradle or Maven integration. * Cross-platform * Lightweight * Optimized: data views, automatic datatypes, SIMD, parallelized batching -* [Neuralang](tutorials/Neuralang.md) scripting language for model definition - -## :rocket: Setup and links - -Add to your dependencies the JAR file of a specific version. Download this -from the project's [releases](https://github.com/MKLab-ITI/JGNN/releases). -Alternatively, include the latest nightly version as a Maven or Gradle dependency -by following the instructions of the JitPack distribution: +* Neuralang scripting language for model definition -[![](https://jitpack.io/v/MKLab-ITI/JGNN.svg)](https://jitpack.io/#MKLab-ITI/JGNN) +Feel free to contribute in any way, for example through the [issue tracker](https://github.com/MKLab-ITI/JGNN/issues). In addition to bug reports, +requests for features and clarifications are welcome. -:cyclone: [Guidebook](https://mklab-iti.github.io/JGNN/) +## :rocket: [Guidebook](https://mklab-iti.github.io/JGNN/) ## :notebook: Citation @@ -45,24 +36,5 @@ by following the instructions of the JitPack distribution: } ``` -
- Changes since the publication's v1.0.0 - -* Introduced [Neuralang](tutorials/Neuralang.md) -* Autosized parameteters -* Up to 30% less memory -* Up to 80% less running time -* Renamed `GCNBuilder` to `FastBuilder` -* Neighbor attention and message passing -* Sort pooling and graph classification - -
- - -## :thumbsup: Contributing - -Feel free to contribute in any way, for example through the [issue tracker](https://github.com/MKLab-ITI/JGNN/issues). In addition to bug reports, -requests for features and clarifications are welcome. -**Copyright Emmanouil Krasanakis (maniospas@hotmail.com), Apache license 2.0** - \ No newline at end of file +**Apache license 2.0. Copyright © 2024, Emmanouil Krasanakis (maniospas@hotmail.com).** \ No newline at end of file diff --git a/docs/index.html b/docs/index.html index 70c150b..fe2455a 100644 --- a/docs/index.html +++ b/docs/index.html @@ -252,7 +252,8 @@

2. Quickstart

notation. Symbolic expressions are part of a scripting-like language, called Neuralang, that is covered in section 4. However, for faster onboarding we stick to the FastBuilder class for creating models. This class's constructor - creates two constants A and h0 from its two arguments to. Other constants + creates two constants A and h0 from its two arguments, respectivel holding + the graph's adjacency matrix and node features. Other constants and input variables can be set too, but more on this later. After instantiation, we use the following model builder methods to constuct a model. Some of these methods parse symbolic expressions to fastly declare machine learning components.