Shashank

A primer on deep learning compiler technologies in PyTorch for graph capture, intermediate representations, operator fusion, and optimized C++ and GPU code generationillustration by authorComputer programming is magical. We write code in human readable languages, and as though by magic, it gets translated into electric currents through silicon transistors making them behave like switches and allowing them to implement complex logic — just so we can enjoy cat videos on the internet. Between the programming language and…

The lifestyle off-roader SUV, Jimny, from the country’s largest automaker, Maruti Suzuki India Limited, is one of the most highly anticipated cars in India at the moment. The entire country is buzzing to hear the official pricing of the vehicle from the company. Recently, Shashank Srivastava, the Executive Director at Maruti Suzuki India Ltd, commented on the same. In his recent interview, the Maruti chief stated that the Jimny would be priced very competitively in the Indian SUV market. During an interview with…

The media reported that health-tech unicorn Innovaccer has sacked nearly 245 employees, or about 15 percent of its workforce, across teams in India and the US. Innovaccer cofounder and CEO Abhinav Shashank cited an "uncertain macroeconomic environment" as the reason behind the job cuts, according to an internal mail sent to employees and accessed by leading startup news portal Inc42."Innovaccer has always believed in empowering its customers to accelerate their transformations, leading to higher quality care for patients…

IntroductionIn 2014, Ian Goodfellow came up with the idea of GAN or vanilla GAN as we call it today. Although impressive, it was notoriously hard to train. Vanilla GAN suffered from an inability to converge, vanishing gradients, and mode collapse. Subsequently, a ton of research centred around addressing these issues. Researchers experimented with different model architectures, loss functions, and training methodologies. A particularly effective solution was Wasserstein GAN, introduced in 2017 by Arjovsky et al.Photo by…

BackgroundI find it hard to keep my hands steady while clicking pictures. Naturally, many of them suffer from a blur. If you relate to this problem, you’re at the right place. While I can’t teach you photography, I can show you a technique to de-blur images with a few lines of python code. Given the craze around deep learning, one solution would be to train an autoencoder. But, they are computationally expensive to train and sometimes, classical methods work just fine. So, we will instead use the Fourier transform.Fourier…

Deep CNNs used for segmentation often suffer from vanishing gradients. Can we combat this by calculating loss at different output levels?Image segmentation entails partitioning image pixels into different classes. Some applications include identifying tumour regions in medical images, separating land and water areas in drone images, etc. Unlike classification, where CNNs output a class probability score vector, segmentation requires CNNs to output an image.Image segmentation of a tennis player (Source-Creative Commons…

Learn about how Docker simplifies access to NVIDIA GPUs, AWS Inferentia and scaling ML containers on Kubernetesillustration by authorIf you told me a few years ago that data scientists would be using Docker containers in their day to day work, I wouldn’t have believed you. As a member of the broader machine learning (ML) community I always considered Docker, Kubernetes, Swarm (remember that?) exotic infrastructure tools for IT/Ops experts. Today it’s a different story, rarely a day goes by when I don’t use a Docker…

Efficient algorithms and methods in machine learning for AI accelerators — NVIDIA GPUs, Intel Habana Gaudi and AWS Training and Inferentiaillustration by authorChances are you’re reading this blog post on a modern computer that’s either on your lap or on the palm of your hand. This modern computer has a central processing unit (CPU) and other dedicated chips for specialized tasks such as graphics, audio processing, networking and sensor fusion and others. These dedicated processors can perform their specialized tasks much…

Photo by Edward Howell on Unsplashfrom a proof of concept to production and in betweenThe journey of a machine learning project is a complex process. It requires a thorough understanding of data science and statistical techniques and the ability to work with research, engineering, and product teams to deliver a stellar machine learning-driven product.In this article, we’ll explore and establish a basic framework for delivering a machine learning project.Project InitiationData ExplorationData ProcessingModel…

Maruti Suzuki – India’s largest car maker – seeks to make major inroads into the SUV segment, which is growing much faster than other car segments in India. Maruti Suzuki has launched the all-new Brezza, and in days from now, will unveil another all-new SUV in the form of the Grand Vitara. The Grand Vitara, which will be unveiled on the 20th July, will be followed by many other SUVs. Maruti Suzuki plans to have a presence in all SUV segments, and this information comes straight from Mr. Shashank Srivastava, a senior…