Home Artificial Intelligence News Introducing Menger, Large-Scale Distributed Reinforcement Learning Infrastructure

Introducing Menger, Large-Scale Distributed Reinforcement Learning Infrastructure

Audio version of the article

Introducing Menger, Large-Scale Distributed Reinforcement Learning Infrastructure

Introducing Menger, Large-Scale Distributed Reinforcement Learning Infrastructure

Reinforcement learning (RL) is a significant area of machine learning, with the potential to solve a lot of real world problems in various fields, like game theorycontrol theoryoperations researchinformation theorysimulation-based optimizationmulti-agent systemsswarm intelligence, and statistics. Reinforcement learning (RL) infrastructure is a loop system of data collection and training using actors as data sample collectors, and learners to train and update the model.

Reinforcement learning (RL) techniques require many iterations over batches of millions of samples from the environment to learn a target task. For example, games like Dota 2 learn from batches of 2 million frames every 2 seconds. So, a Reinforcement Learning (RL) infrastructure not only demands an efficient increase in the number of actors and samples but also needs smooth and swift iterations over a massive number of samples during the training process.

Introducing Menger, Large-Scale Distributed Reinforcement Learning Infrastructure

Overview of a Reinforcement Learning (RL) system in which an actor sends trajectories (e.g., multiple samples) to a learner. The learner trains a model using the sampled data and pushes the updated model back to the actor (e.g., TF-AgentsIMPALA).

Thus, to serve this demand, Google AI introduces Menger, which is a massive large-scale distributed RL infrastructure. This reduces the overall training time by up to 8.6x compared to a baseline implementation, thus increasing the efficiency of iterations. Menger is implemented using Google TPU accelerators for fast training iterations. Menger uses local inference (rather than a centralized one) but pushes actors’ scalability to virtually an unbounded limit. This plays a major role in dealing with the complicated task of chip placement.

Menger is expected not only to further progress in the chip design process but also in other sophisticated real-world tasks as well.

This article has been published from a wire agency feed without modifications to the text. Only the headline has beeen changed.

Source link

- Advertisment - Introducing Menger, Large-Scale Distributed Reinforcement Learning InfrastructureIntroducing Menger, Large-Scale Distributed Reinforcement Learning Infrastructure

Most Popular

SwiftUI TabView Introduction and Tab Bar Customization

  The tab bar interface appears in some of the most popular mobile apps such as Facebook, Instagram, and Twitter. A tab bar appears at...

Preparing Data in Machine Learning

Data preparation may be one of the most difficult steps in any machine learning project. The reason is that each dataset is different and highly...

Using blockchain to crack down abusive imagery

Blockchain could be an effective and efficient solution for helping to rid the internet of abusive imagery. Tackling abusive imagery can help victims...

Transforming chatbots throgh AI and ML

The increasing technology has always been a saviour for us. Technology still provides us with solutions for existing problems. One of the answers offered...

Some tricks in Python

Python is one of the most popular programming languages ​​for beginner developers, making it the most widely taught language in schools around the world. However,...

Introduction to R Markdown

In this blog post, we’ll look at how to use R Markdown. By the end, you’ll have the skills you need to produce a...
- Advertisment - Introducing Menger, Large-Scale Distributed Reinforcement Learning InfrastructureIntroducing Menger, Large-Scale Distributed Reinforcement Learning Infrastructure