In 2012 I received my PhD in experimental condensed matter physics from West Virginia University. Following that I was a dry etch process engineer for Intel Corporation, where I leveraged big data to make essential process improvements for mission critical products. After leaving Intel in 2015, I have worked as a contract and freelance deep learning and artificial intelligence engineer.

Leave a Reply to katia010601 Cancel reply

katia010601 says:

April 4, 2022 at 3:26 pm

Hello,
I am trying to work on a similar implementation of this and I really need to set the seeds for reproducibility. I have tried everything but it does not work as expected. I have tried setting seeds for torch, random , numpy , the environment and the environment action space both in the main and worker function but no matter what i cannot achieve reproducibility. Would you have any insight on how I could achieve it ?

1. phil says:
  
  April 4, 2022 at 3:33 pm
  
  The easiest way is to drop a link to your code on github and I can take a look.
  
  Just as a general strategy, what I try to do in these situations is strip out complexity. Start with a very basic script where you just generate random numbers / reset the environment. Make sure you can get reproducibality there, and then compare to the more complex code.
  
  1. katia010601 says:
    
    April 5, 2022 at 8:41 am
    
    Thank you for the reply.
    Yes I might try doing in a more simple implementation, but the problem i think lies in the shared memory remember function or when loading the model with state_dict .
    I tried fixing the seeds on your simple cartpole implementation without the encoders but I couldnt even do it there.
    
    This is my repository https://github.com/katiavas/rep.git
    
  2. lamare34 says:
    
    December 12, 2022 at 11:02 am
    
    Hello Mr.Tabor,
    
    I want to save the work and test it in different environments to see its performance. This is one of the missing point in your tutorial. Can you share this situation with the necessary additions? I tried a lot but couldn't find the solution.

Curiosity Driven Deep Reinforcement Learning

Ready?

Curiosity Driven Deep Reinforcement Learning

What You Will Learn in This Course

How to Succeed in This Course

Required Background, Software, and Hardware

A Brief Review of Deep Reinforcement Learning and Actor Critic Methods

Code Review of Basic Actor Critic Agent

A Crash Course in Asynchronous Advantage Actor Critic Methods

Our Code Structure

How to Read and Implement Research Papers

A3C Paper: Abstract and Introduction

Crash Course in Parallel Processing in Python

A3C Paper: Related Work, Reinforcement Learning Background

A3C Paper: The Asynchronous Reinforcement Learning Framework

Coding our Actor Critic Network

Learning with Generalized Advantage Estimation

Coding a Minimalist Replay Memory

Coding the Shared Adam Optimizer

A3C Paper: Experiments and Discussion

How to Modify the Open AI Gym Atari Environments

Coding Our Main Loop and Evaluating Our Agent

Paper Overview

ICM Paper: Abstract and Introduction

ICM Paper: Curiosity Driven Exploration

Experimental Setup and Coding Our ICM Module

ICM Paper: Experiments, Related Work, and Discussion

Setting Up the Mini World and Training Our ICM Agent

Making our Virtual Environment

Making our Agent Compliant with the New Gym Interface

About the teacher

phil