PsychRNN
v1.0.0-alpha

Contents:

  • Installation Guide
  • API Documentation
  • Getting Started
    • Hello World!
    • Simple Example
    • Biological Constraints
    • Curriculum Learning
    • Accessing and Modifying Weights
    • Loading Model with Weights
    • Simulation in NumPy
    • Define New Task
    • Define New Model
    • Further Extensibility – Initializations, Loss Functions, and Regularizations
PsychRNN
  • Docs »
  • Getting Started
  • Edit on GitHub

Getting Started¶

Each guide below includes a link to a Colab notebook that will let you experiment with each example on your own in the browser.

Contents:

  • Hello World!
  • Simple Example
    • Initialize Task
    • Initialize Model
      • Set Up Network Parameters
        • Initialization Parameters
        • Regularization Parameters
      • Instantiate Model
    • Train Model
      • Set Up Training Parameters
      • Train Model on Task using Training Parameters
    • Test Model
    • Get & Save Model Weights
    • Cleanup
  • Biological Constraints
    • Biologically Unconstrained
    • Dale Ratio
    • Autapses
    • Connectivity
    • Fixed Weights
  • Curriculum Learning
    • Instantiate Curriculum Object
    • Initialize Models
    • Train Models
      • Plot Losses
    • Cleanup
  • Accessing and Modifying Weights
  • Loading Model with Weights
    • Load from File
    • Load from Weights Dictionary
  • Simulation in NumPy
    • Load from Model
      • Simulate Model
    • Load from File
      • Simulate Model
    • Load from Dictionary
      • Simulate Model
  • Define New Task
  • Define New Model
    • init
    • forward_pass
      • output_timestep
      • recurrent_timestep
  • Further Extensibility – Initializations, Loss Functions, and Regularizations
Next Previous

© Copyright 2020, Daniel B. Ehrlich*, Jasmine T. Stone*, David Brandfonbrener, Alex Atanasov, John D. Murray (* indicates equal contribution) Revision 380ee812.

Built with Sphinx using a theme provided by Read the Docs.