MLOps Hype, Open Source Risks & AI Research Fallacies

The Belamy | Weekly dose of best AI stories

Jul 18

As we come out of lockdowns, the economic activity seems to pick up aggressively, data science hiring has increased; so has the inventory of our upcoming events.

Here's the list of online events we have lined up for the next 2 months. Go register.

Webinar | How To Get Started With A Career In Data Science | 27th Jul 2021

Hands-on Workshop | Accelerate PyTorch Applications Using Intel oneAPI Toolkit | 18th Aug 2021

Virtual Conference | Rakuten Applied AI Conference 2021 | 19-20th Aug 2021

Virtual Conference | Deep Learning DevCon 2021 | 23-24th Sep 2021

Does MLOps Live Upto The Hype?

Deeplearning.ai recently hosted a panel of MLOps experts to derive insights on the most important aspects of production machine learning and what MLOps looks like at companies. Hosted by Ryan Keenan of Deeplearning.ai, the panel consisted of Andrew Ng, Robert Crowe, Lawrence Moroney, Chip Huyen and Rajat Monga.

The panel of experts began by addressing the significance of MLOps in today’s world. Chip Huyen, who teaches ML at Stanford, considers model training to be a small part of the problem. According to her, the problem is retraining. Once the model is out in the open, data drifts can happen. So how does one keep on updating and compensating for these variations?

Security Risks Of Open Source Software

Be it Linux or Tensorflow, the open source community plays a huge role in taking a cutting edge technology mainstream. For instance, Tensorflow accelerated the popularity of machine learning after it was open sourced in 2015.

However, open source comes at a cost. Due to the involvement of third parties/developers, open source projects are prone to vulnerabilities. And when the components of these open source projects show up as libraries and kernels in mega projects (think: self-driving cars), the outcomes can be irreversible.

Four Fallacies In AI Research

Research in AI often follows a cyclic pattern: periods of rapid progress, successful commercialisation, heavy public and private investments, called AI Spring, is often followed by AI winter, characterised by waning enthusiasm, drying up of funding and jobs.

Over-optimism among people, the media and even experts arise from fallacies in our understanding of AI and the intuitions about the nature of intelligence.

JAX Vs TensorFlow Vs PyTorch

Popular libraries such as TensorFlow and PyTorch keep track of gradients over neural network parameters during training with both comprising high-level APIs for implementing the commonly used neural network functionality for deep learning.

JAX is NumPy on the CPU, GPU, and TPU, with great automatic differentiation for high-performance machine learning research. Along with a Deep Learning framework, JAX has created a super polished linear algebra library with automatic differentiation and XLA support.

How Reinforcement Learning Is Advancing Chip Designing

Arranging ‘billions’ of components on a tiny surface area of a computer chip is a complicated process. It calls for precise decision-making at every step of the way and requires a designer with years of experience in laying out circuits that squeeze power efficiency from nanoscopic devices.

Designers now tap the latest AI advancements to learn the processes involved in chip designing to help draw up more powerful blueprints in less time. It allows engineers to co-design an AI software to find the optimal configuration with different designer perspectives.

Hands-on Guides for ML Developers

+ Complete Guide To Bidirectional LSTM (With Python Codes)

+ How to do Pose Estimation With MoveNet

+ How To Do Image Segmentation Using DeepLab?

+ Beginner’s Guide To Machine Learning With Apache Spark

+ Guide To VOLO: Vision Outlooker For Visual Recognition

PEOPLE & STARTUPS

+ Krish Naik Speaks About His ML Journey & Advice To Data Scientists

+ Bayer Pharmaceutical’s Abhishek Choudhary Traces His Machine Learning Journey

+ Interview: Avinash Gupta, Managing Director & CEO - India, Dun & Bradstreet

+ Every Good CISO Comes With A Healthy Dose Of Professional Paranoia: Jaya Baloo, Avast Antivirus

EleutherAI’s GPT-J vs OpenAI’s GPT-3

OpenAI’s not so open GPT-3 has an open-source cousin GPT-J, from the house of EleutherAI. EleutherAI, founded by Connor Leahy, Leo Gao, and Sid Black, is a research group focused on AI alignment, scaling and open-source AI research. In March 2021, the company released two GPT-Neo models with 1.3 billion and 2.7 billion parameters respectively.

Google's Second Cloud Region In India

Google launched its second cloud region in Delhi NCR, India. With its latest launch, Google now has 26 cloud regions, 79 zones and 146 points of presence worldwide. In the Asia Pacific, Delhi NCR marks the 10th region. In India, the first cloud region was set up in Mumbai in 2017.

The Delhi and Mumbai regions will enable a geographically separate in-country network for customers’ mission-critical applications.

BOTTOM OF THE NEWS

Here's what all happened last week.

+ IIT Madras’ Team Avishkar Qualifies For European Hyperloop Week

+ Fleet Safety Solution Provider Netradyne Raises $150 Mn

+ Autonomous Industrial Vehicle Maker Ati Motors Raises $3.5 Mn

+ AWS Public Sector Startup Ramp Launched In India

+ IIT Delhi Establishes A Chair In Data Analytics To Attract Best Talents