Engineering @ TCS | Prev @ Deloitte | IBM Quantum Qiskit Advocate | Independent Researcher

github

medium

projects

Soumyadip Sarkar • Peerlist

Minitensor is a lightweight, high-performance tensor operations and automatic differentiation library designed for developers seeking a PyTorch-like experience with enhanced speed and efficiency. Built with a Rust backend, it ensures maximum speed and memory efficiency, while its Python bindings provide a familiar API for easy adoption. Minitensor supports complete neural network layers and optimizers, offering seamless interoperability with NumPy arrays. Key features include automatic differentiation for gradient computation, a modular design for customization, and a high-performance engine optimized with SIMD. The library is currently in active development, making it a cutting-edge choice for developers looking to build and deploy intelligent applications.

MiniTensor

Test your reflexes and number recognition skills by popping the right bubbles before time runs out!

Bubble Game

Easily design, simulate, and visualize quantum circuits with intuitive tools that allow you to experiment with quantum gates and qubit states effortlessly.