alexlenail/NN-SVG
NN SVG
Publication-ready NN-architecture schematics.
Features
- Neural Network Diagram Generators - Creates publication-ready SVG schematics of neural network architectures with customizable layer configurations.
- Neural Architecture Visualizations - Creates publication-ready SVG diagrams of neural network architectures for academic papers and presentations.
- Convolutional Neural Network Architectures - Creates publication-ready SVG schematics of CNN architectures with configurable convolutional and pooling layers.
- AlexNet Diagram Generation - Generates publication-ready SVG diagrams of AlexNet-style architectures with parallel GPU paths.
- AlexNet Diagram Renderers - Generates publication-ready SVG schematics of AlexNet-like networks with parallel GPU paths.
- CNN Diagram Generation - Generates publication-ready SVG diagrams of convolutional neural networks showing feature maps and filters.
- Fully Connected Network Diagram Generation - Generates publication-ready SVG diagrams of fully connected neural networks with configurable layer sizes.
- Fully Connected Network Diagrammers - Draws classic feedforward neural network diagrams with configurable layer sizes and connections.
- CNN Diagram Generators - Produces convolutional neural network schematics showing feature maps, filters, and pooling layers.
- LeNet and AlexNet Layout Templates - Ships pre-built layout algorithms for generating diagrams of LeNet and AlexNet-style architectures.
- Convolutional Network Schematics - Produces detailed CNN schematics showing feature maps, filters, and pooling operations for research documentation.
- SVG Neural Network Diagrammers - Creates scalable vector graphics of neural network topologies for academic papers and presentations.
- Neural Network Layer Configurations - Ships a declarative JSON configuration system for defining network layer types, dimensions, and visual properties.
- AlexNet Diagram Generators - Creates publication-ready SVG schematics of AlexNet-like architectures with parallel GPU paths.
- Neural Network Diagram Coordinate Grids - Provides the coordinate grid system that positions all tensors and filters in generated neural network diagrams.
- SVG Export Pipelines - Serializes the in-memory SVG DOM tree to a standalone file string for download without external dependencies.
- D3.js Implementations - Uses D3.js to render all neural network diagrams by manipulating SVG elements through data-binding.
- Random Weight Initialization for Diagrams - Assigns random values to connection weights and filter kernels to produce parametric diagram visualizations.
- Neural Network Diagram Styling - Provides adjustable visual properties for neural network diagrams including colors, opacity, and spacing.
- Neural Network Dimension Annotations - Adds tensor and convolutional filter dimension labels directly onto neural network architecture diagrams.
- Network Topology Diagrams - Defines fully-connected, convolutional, or AlexNet-style topologies with random weight initialization for parametric diagrams.
- SVG Document Exporters - Downloads the current network schematic as a standalone SVG file for use in publications and presentations.
- SVG Image Exports - Saves neural network architecture drawings as Scalable Vector Graphics files for publications or web pages.
- Neural Network SVG Exporters - Saves neural network architecture drawings as scalable vector graphics for publications and presentations.
- Diagram Customization - Adjusts visual properties including node size, colors, edge width, opacity, and layer spacing for tailored schematics.
- Neural Network Visual Property Cascades - Merges global default styles with per-layer overrides for colors, opacity, and node dimensions in diagrams.
