IGNNITION at a glance

Why IGNNITION?

In recent years Graph Neural Networks (GNNs) have proven to be a very effective tool to design extremely successful models. The main disadvantage of its use is the high complexity of implementing the actual model as for many scientists/practitioners, implementing a model using native ML frameworks (e.g., Tensorflow, PyTorch) would require weeks or even months.

For this reason, several frameworks have been developed to help this users in this task (e.g., Graph Nets, DGL, Spektral). However, as far as we know, non of them ultimately achieve the goal since they either lack of sufficient flexibility, or they require the implementation of several parts of the algorithm.

IGNNITION is the first framework to overcome these difficulties and provide a framework that allows very fast prototyping of GNNs without any knowledge of ML frameworks, and with exceptional flexibility, to adapt to any possible design of GNNs.

Main functionalities:

High-level abstraction

Designing a GNN can be very challenging, especially in complex problems were conceptualizing the GNN is a non-trivial task. For this reason, we introduce a novel abstraction technique, which we named Multi-Stage Message Passing (MSMP), which is intended to help simpy and conceptualize a GNN model. The main advantage of the use of this abstraction is that it represents visually the main parts of the GNN (helping to gain a visual intuition) meanwhile, no major assumptions are made. Hence, our abstraction is general enough to adapt to any GNN design.

No coding is needed

IGNNITION is intended to be an extremely simple tool so that it is accessible to non-ML experts. In the image below we can observe how the user must only define the model description file using a declarative language (YAML) as well as to provide the dataset. Then, with three simple lines of code, the user can call the core engine of IGNNITION that will run the desired functionality and ultimately produce the trained GNN.

workflow

workflow

The image below shows an example of how the IGNNITION framework can be called from a python script to start the training of a GNN.

// main.py
import ignnition

def main():
    model = ignnition.create_model(model_dir = <PATH>)
    model.train_and_validate()

High flexibility

To date, all existing frameworks for fast prototyping of GNNs either require the actual coding of the model or lack enough flexibility to support the vast range of possible GNN models. IGNNITION overcomes this issue as even though no Tensorflow code is needed, this does not incur any flexibility loss. Consequently, potentially any GNN model can be implemented using our framework.

Easy debugging

One of the most difficult tasks that a Machine Learning Engineer faces is the debugging phase. ML models act as black-boxes, which make it hard to understand their inner working. For this reason, IGNNITION incorporates a debugging assistant that will help users to easily identify possible malfunctions, and even suggest possible fixes for most errors in the definition of the model.

Easy integration

Traditionally all ML models are strongly dependent on the dataset it is applied to, which implies that several non-trivial adaptations of the models are needed to be able to reuse it for a different dataset. To solve this problem, the definition of a GNN in IGNNITION is agnostic to the dataset, it will be applied to, for which it can easily be integrated to any available dataset.

High performance

Even though IGNNITION allows its users to isolate entirely from the actual implementation of the GNN, this does not incur in any major overhead time-wise. Hence, our implementations result to be essentially as efficient as an actual native Tensorflow implementation.

Next step

IGNNITION is intended to be useful for both inexperienced and expert users in the field of GNNs. Nevertheless, we recommend the inexperienced user to refer to the section GNN background which provides a basic intuition that can help you speed up the design of your first GNN! For those with already some experience, we recommend proceeding with the installation guide to start using IGNNITION.