DocsEngine

CyxWiz Engine

The flagship desktop application for designing, training, and deploying machine learning models with a visual interface similar to professional game engines.

Overview

The Engine combines:

Visual Node Editor

Drag-and-drop ML pipeline building

Python Integration

Full scripting support with embedded interpreter

Real-Time Visualization

Live training metrics and plotting

70+ Tool Panels

Comprehensive data science toolkit

Code Generation

Export to PyTorch, TensorFlow, Keras

Main Interface

+------------------------------------------------------------------+
|  File  Edit  View  Nodes  Train  Dataset  Script  Plots  Deploy  |
+------------------------------------------------------------------+
|                    |                            |                 |
|   Asset Browser    |     Node Editor            |   Properties    |
|                    |                            |                 |
|   [Project Files]  |   [Visual Graph]           |   [Selected     |
|   - Scripts/       |                            |    Node Params] |
|   - Models/        |   +------+    +-------+    |                 |
|   - Datasets/      |   |Input |---->|Dense  |   |   Units: 128    |
|                    |   +------+    +-------+    |   Activation:   |
|                    |                   |        |   [ReLU      v] |
|                    |               +-------+    |                 |
+--------------------+               |Output |    +-----------------+
|                                    +-------+                      |
|   Console / Script Editor / Training Dashboard                    |
|                                                                   |
|   >>> import pycyxwiz                                             |
|   >>> model = pycyxwiz.Sequential()                               |
|                                                                   |
+------------------------------------------------------------------+

Documentation Sections

Core Features

Section	Description
Menus Reference	Complete menu bar documentation
Keyboard Shortcuts	All shortcuts and customization
Command Palette	Quick command access (Ctrl+P)

Panels

Panel	Description
Node Editor	Visual ML pipeline builder
Script Editor	Python code editing
Console	Python REPL and logs
Properties	Node parameter editing
Asset Browser	Project file management
Training Dashboard	Real-time loss/accuracy plots
Dataset Panel	Dataset configuration

Feature Highlights

Visual Model Building

Build neural networks visually without writing code:

80+ Node Types - Layers, activations, optimizers, loss functions
Smart Connections - Type-safe pin connections with validation
Shape Inference - Automatic output shape calculation
Code Export - Generate PyTorch, TensorFlow, Keras, or PyCyxWiz

Comprehensive Tool Suite

Access 70+ analysis tools via the Tools menu:

Data Science - Profiling, correlation, missing values
Statistics - Hypothesis testing, regression analysis
Clustering - K-Means, DBSCAN, GMM, hierarchical
Model Evaluation - ROC curves, confusion matrices

Signal Processing - FFT, wavelets, spectrograms
Linear Algebra - SVD, eigendecomposition, QR
Time Series - Forecasting, decomposition, seasonality
Text Processing - Tokenization, TF-IDF, embeddings

Real-Time Training

Monitor training progress live:

Loss/Accuracy Plots - Updated every batch
Live Metrics - Current epoch, learning rate
GPU Utilization - Memory and compute usage
Export Data - Save training history to CSV

System Requirements

Minimum

OS: Windows 10, macOS 10.15, Ubuntu 18.04
CPU: 4 cores
RAM: 8 GB
GPU: OpenGL 3.3 compatible
Storage: 500 MB (application)

Recommended

OS: Windows 11, macOS 13, Ubuntu 22.04
CPU: 8+ cores
RAM: 16+ GB
GPU: NVIDIA RTX series (for CUDA acceleration)
Storage: 10+ GB (with datasets)

File Types

Extension	Description
.cyxgraph	Node editor graph files
.cyxmodel	Trained model files
.cyx	CyxWiz script files
.py	Python script files