Home

Welcome to the TensorMap wiki!

Tensormap is a web application that enables users to create deep learning models using a graphical interface without having to know how to code. We hope that this will help as a stepping stone to individuals who are starting to explore the deep learning paradigm.

Listed below are the main functions of Tensormap

As a part of Google Summer of Code 2019 we implemented the following features:

• Create neural network architecture using drag and drop interface.
  • Create different nodes(input, hidden, output) by dropping to the workspace.
  • Link different nodes that define flow to data and model.
  • Group Various node to form a layer.
  • Define different parameters associated with nodes and layers.
  • Manually update and retrieve weight of links.
  • specify model compilation and execution configuration (like learning rate, optimizer, etc)
  • View Runtime results.
  • Get the generated code.
• Upload data of CSV format and visualize the uploaded data
• Perform data manipulations on uploaded data like:
  • Adding rows
  • Editing rows
  • Deleting rows
  • Deleting columns
  • Sorting columns
  • Filtering column data
  • Searching for data
• Specify dataset and experiment related configurations like:
  • Features
  • Labels
  • Test percentage
  • Experiment type (Binary Classification, Multiclass Classification, Regression)
  • Number of epochs
  • Batch size
  • Loss function (Binary Crossentropy, Categorical Crossentropy, Mean Squared Error )
  • Optimizer (Adam optimization)
• Download edited CSV
• Download the code that was auto-generated according to the created model
• Execute the created model and show the progress
• Show the resultant metric values