This is your work, valued

Cambridge

Riccardo Grazzi

Advanced
@prolearner

Researcher at MSR Cambridge, previously at IIT in Genoa, Italy. Working on principled, efficient optimization for machine learning and on LLMs' expressivity.

prolearner.github.io/riccardograzzi.github.io

hypertorch. Python

126

procedural-planet. a 3D procedural Planet in THREEJS and WebGl

49

hyper-representation. This is the official repo for the experiments in the paper "Bilevel Programming for Hyperparameter Optimization and Meta-Learning"

29

LearningToCompareTF. implementation of the learning to compare article using tensorflow

11

onlineLTL. python implementation of online learning to learn non-smooth algorithms

3

nonsmooth_implicit_diff. Code for the paper Nonsmooth Implicit Differentiation: Deterministic and Stochastic Convergence Rates by Riccardo Grazzi, Massimiliano Pontil and Saverio Salzo (ICML 2024).

2