BrowserGNN Training Dashboard

Real-time Graph Neural Network Training in Your Browser

Dataset

Karate Club

34 nodes, 2 classes

Social

20 nodes, 2 classes

Citation

25 nodes, 3 classes

Molecule

16 nodes, 3 classes

Protein

30 nodes, 3 classes

Web Graph

25 nodes, 3 classes

Custom CSV

Upload your own

📁

Click to upload or drag & drop
CSV files (edges, features, labels)

Format: edges.csv (src,dst), features.csv (one row per node), labels.csv (class per node)

custom-dataset

Train/Val Split 70% train / 30% val

Model Architecture

GNN Layer Type

Hidden Channels

Number of Layers

Dropout

Training Parameters

Optimizer

Learning Rate

Weight Decay

Max Epochs

Early Stopping Patience

Training Control

Ready to train

Current Metrics

Epoch

Loss

Train Acc

Val Acc

Loss Curves

Train Loss

Val Loss

Accuracy Curves

Train Accuracy

Val Accuracy

Graph Visualization

Training Log

[System] BrowserGNN Training Dashboard initialized

[System] Select a dataset and click "Start Training" to begin