• Collaborating in a team of 4 to build software for MelonTech, a University of Manchester research start-up that is developing tools to help patients with Congenital Hyperinsulinism (CHI).

• Converted the research prototypes created by the academic team into a scalable and secure web service that was sufficiently robust to support wider scale medical trials with the technology.

• Put in place appropriate data security and data protection mechanisms, to ensure that personal and sensitive data was safeguarded.

My responsibilities included engaging with academics and ML practitioners by creating content (open source software, in-depth paper summaries, reproducing research papers, community outreach) and collaborating across internal teams to improve our integrations with popular deep learning frameworks.

Weights and Biases (W&B) Authors Program is an initiative run by the company to make Deep Learning and Machine Learning topics accessible via reports. As a Weights and Biases Author, I am responsible for the following:

1. Write reports on the state of the art in Deep Learning research.
2. Make research papers accessible to the world.
3. Suggest product feedback.

Open Source Contributor

Open Source Contributor

Open Source Contributor

This article provides an overview of "Adaptive Budget Allocation for Parameter Efficient Fine-Tuning" using W&B for interactive visualizations. It includes code samples for you to follow!

This article provides an overview of "QLoRA: Efficient Finetuning of Quantized LLMs" using W&B for interactive visualizations. It includes code samples for you to follow!

This article givens an overview of "Low-Rank Adaptation of Large Language Models" using W&B for interactive visualizations. It includes code samples for you to follow.

This article provides a brief overview of intrinsic dimensions and how they enable Low-Rank Domain Adaptation. We also provide code samples which use Weights & Biases for interactive visualizations.

This article provides an overview of the Mixture Model Networks (MoNet) architecture, with code examples in PyTorch Geometric and interactive visualizations using W&B.

This article provides a brief overview of the Residual Gated Graph Convolutional Network architecture, complete with code examples in PyTorch Geometric and interactive visualizations using W&B.

This article provides a brief overview of the Graph Attention Networks architecture, complete with code examples in PyTorch Geometric and interactive visualizations using W&B.

This article provides an overview of the GraphSAGE neural network architecture, complete with code examples in PyTorch Geometric, and visualizations using W&B.

This article provides a beginner-friendly introduction to Convolutional Graph Neural Networks (GCNs), which apply deep learning paradigms to graphical data.

This article provides a beginner-friendly introduction to Attention based Graphical Neural Networks (GATs), which apply deep learning paradigms to graphical data.

This article provides a beginner-friendly introduction to Message Passing Graph Neural Networks (MPGNNs), which apply deep learning paradigms to graphical data.

Breakdown of Emerging Properties in Self-Supervised Vision Transformers by Mathilde Caron, Hugo Touvron, Ishan Misra, Hervé Jégou, Julien Mairal, Piotr Bojanowski and Armand Joulin with Weights and Biases logging.

An in-depth breakdown of "Graph Neural Networks with Learnable Structural and Positional Representations" by Vijay Prakash Dwivedi, Anh Tuan Luu, Thomas Laurent, Yoshua Bengio and Xavier Bresson.

In depth analysis of a pure MLP based architectures, viz. "FNet: Mixing Tokens with Fourier Transforms" by James Lee-Thorp, Joshua Ainslie, Ilya Eckstein and Santiago Ontañón.

In the past few months there have been various papers proposing MLP based architectures without Attention or Convolutions. This report analyses the paper 'ResMLP: Feedforward networks for image classification with data-efficient training' by Touvron et al.

It's no news that transformers have dominated the field of deep learning ever since 2017. But in their recent work, titled 'Pay Attention to MLPs,' Hanxiao Liu et al. propose a new architecture that performs as well as Transformers in key language and vision applications. Let's dig in.

In the past few months there have been various papers proposing MLP based architectures without Attention or Convolutions. This report analyses the paper 'MLP-Mixer: An all-MLP Architecture for Vision' by Ilya Tolstikhin, Neil Houlsby, Alexander Kolesnikov, Lucas Beyer and others.

Pruning can be a clever way to reduce a model's resource greediness. But what gets forgotten when you do?