Topic Modeling Neural Toolkit Documentation

The Topic Modeling Neural Toolkit (TMNT) provides implementations for training and applying neural network variational autoencoders (VAEs) on text data in order to model latent topics in large documents collections.

Table of Contents

Overview

• Installation
  • Environment
  • Pip Installation
  • Source Installation
• TMNT Summary of Capabilities
• Common Formats
  • Input Documents
  • Configuration Files
  • Model Space Files

Core

• Quickstart
  • Preprocessing
  • Training a Topic Model
  • Using the Model for Inference
• Evaluation and Visualization
  • Evaluation
  • Visualization
• Inference with Topic Models
  • Document Encodings

Examples

• TMNT Examples

Advanced Use

• Modeling Covariates

API Reference

• PreProcessing
• Modeling
• Inference

Indices and tables

• Index
• Module Index
• Search Page