Popular GitHub Repositories for Lipschitz Constant

RWKV (pronounced RwaKuv) is an RNN with great LLM performance, which can also be directly trained like a GPT transformer (parallelizable). We are at RWKV-7 "Goose". So it's combining the best of RNN and transformer - great performance, linear time, constant space (no kv-cache), fast training, infinite ctx_len, and free sentence embedding.

Translate darknet to tensorflow. Load trained weights, retrain/fine-tune using tensorflow, export constant graph def to mobile devices

Extend existing LLMs way beyond the original training length with constant memory usage, without retraining

PyXAB - A Python Library for X-Armed Bandit and Online Blackbox Optimization Algorithms

Unsupervised multilabel image segmentation (color/gray/multichannel) based on the Potts model (aka piecewise constant Mumford-Shah model)

LipSDP - Lipschitz Estimation for Neural Networks

Constrained deep learning is an advanced approach to training deep neural networks by incorporating domain-specific constraints into the learning process.

Unofficial PyTorch implementation of "GAN-QP: A Novel GAN Framework without Gradient Vanishing and Lipschitz Constraint"

Reference implementation of "Softmax Attention with Constant Cost per Token" (Heinsen, 2024)

A very simple tool to rewrite parameters such as attributes and constants for OPs in ONNX models. Simple Attribute and Constant Modifier for ONNX.

Source code for "Training Generative Adversarial Networks Via Turing Test".

Supplementary material for IEEE Services Computing paper 'An SRAM Optimized Approach for Constant Memory Consumption and Ultra-fast Execution of ML Classifiers on TinyML Hardware'

Code for AAAI 2024 paper: LRS: Enhancing Adversarial Transferability through Lipschitz Regularized Surrogate

Implementation of "Incorporating Learnable Membrane Time Constant to Enhance Learning of Spiking Neural Networks" by Fang et al. (2020)

Diacritics are short vowels with a constant length that are spoken. The same word in the Arabic language can have different meanings and different pronunciations based on how it is diacritized. In this project, we implement a pipeline to predict the diacritic of each character in an Arabic text using Natural Language Processing techniques.

Leveraging network motifs to improve artificial neural networks