JonasGeiping

breaching

Breaching privacy in federated learning scenarios for vision and text

carving

Package to optimize Adversarial Attacks against (Large) Language Models with Varied Objectives

cramming

Cramming the training of a (BERT-type) language model into limited compute.

data-poisoning

Implementations of data poisoning attacks against neural networks and related defenses.

dataaugs

fullbatchtraining

Training vision models with full-batch gradient descent and regularization

invertinggradients

Algorithms to recover input data from their gradient signal through a neural network

linear_cross_entropy_loss

A fusion of a linear layer and a cross entropy loss, written for pytorch in triton.

poisoning-gradient-matching

Witches' Brew: Industrial Scale Data Poisoning via Gradient Matching

torchlmdb

LMDB wrapper for pytorch datasets.

JonasGeiping/cramming

Cramming the training of a (BERT-type) language model into limited compute.

JonasGeiping/invertinggradients

Algorithms to recover input data from their gradient signal through a neural network

JonasGeiping/breaching

Breaching privacy in federated learning scenarios for vision and text

JonasGeiping/poisoning-gradient-matching

Witches' Brew: Industrial Scale Data Poisoning via Gradient Matching

JonasGeiping/data-poisoning

Implementations of data poisoning attacks against neural networks and related defenses.

JonasGeiping/carving

Package to optimize Adversarial Attacks against (Large) Language Models with Varied Objectives

JonasGeiping/linear_cross_entropy_loss

A fusion of a linear layer and a cross entropy loss, written for pytorch in triton.

JonasGeiping/fullbatchtraining

Training vision models with full-batch gradient descent and regularization

JonasGeiping/dataaugs

JonasGeiping/torchlmdb

LMDB wrapper for pytorch datasets.

JonasGeiping/ParametricMajorization

New training strategies for energy minimization methods with learnable parameters (ICCV 2019)

JonasGeiping/poisoning-benchmark

A unified benchmark problem for data poisoning attacks

JonasGeiping/flash-attention

Fast and memory-efficient exact attention

JonasGeiping/flexBox

FlexBox is a fexible MATLAB toolbox for finite dimensional convex variational problems in image processing and beyond.

JonasGeiping/dleval

New and fancy exercises evaluation system for the "Deep Learning" course!

JonasGeiping/gpt-neox

An implementation of model parallel autoregressive transformers on GPUs, based on the DeepSpeed library.

JonasGeiping/jonasgeiping.github.io

JonasGeiping/keras-applications

Reference implementations of popular deep learning models.

JonasGeiping/lm-watermarking

JonasGeiping/meta-shampoo

For optimization algorithm research and development.

JonasGeiping/prost

JonasGeiping/webpage

JonasGeiping/website

📝 Easily create a beautiful website using Hugo, GitHub, and Netlify