Use PEFT or Full-parameter to finetune 350+ LLMs or 100+ MLLMs. (LLM: Qwen2.5, Llama3.2, GLM4, Internlm2.5, Yi1.5, Mistral, Baichuan2, DeepSeek, Gemma2, ...; MLLM: Qwen2-VL, Qwen2-Audio, Llama3.2-V…

[NeurIPS 2019] Deep Leakage From Gradients

The generator of Hu editor's style.

Federated learning with text DNNs for DATA 591 at University of Washington.

The original code for the paper "Learning to Learn via Self-Critique".

Evil Huawei - 华为作过的恶

GNU Radio – the Free and Open Software Radio Ecosystem

Code for the paper "Learning to Demodulate from Few Pilots via Offline and Online Meta-Learning"

Code for the paper "Meta-Learning to Communicate: Fast End-to-End Training for Fading Channels"

Neural Joint-Source Channel Coding

lookahead optimizer (Lookahead Optimizer: k steps forward, 1 step back) for pytorch

Differentiable ODE solvers with full GPU support and O(1)-memory backpropagation.

Meta Learning for Semi-Supervised Few-Shot Classification

Unsupervised Data Augmentation (UDA)

Code for "Turbo Autoencoder: Deep learning based channel code for point-to-point communication channels" NeurIPS 2019

[NOT OFFICIAL VERSION] Communication Algorithms via Deep Learning. Paper: https://arxiv.org/abs/1805.09317

Python package built to ease deep learning on graph, on top of existing DL frameworks.

Binarized Neural Network (BNN) for pytorch

Pytorch implementation of paper "HiDDeN: Hiding Data With Deep Networks" by Jiren Zhu, Russell Kaplan, Justin Johnson, and Li Fei-Fei

A PyTorch Implementation of Federated Learning https://doi.org/10.5281/zenodo.4321561

Ray is a unified framework for scaling AI and Python applications. Ray consists of a core distributed runtime and a set of AI Libraries for accelerating ML workloads.

学习强国 懒人刷分工具 自动学习

Practical correction and capacity estimation of deletion channel using correction trees

Simulates pruned DFT spread FBMC and compares the performance to OFDM, SC-FDMA and conventional FBMC. The included classes (QAM, DoublySelectiveChannel, OFDM, FBMC) can be reused in other projects.

SteganoGAN is a tool for creating steganographic images using adversarial training.

Implementation A Style-Based Generator Architecture for Generative Adversarial Networks in PyTorch

BOAssembler: a Bayesian Optimization Framework to Improve RNA-Seq Assembly Performance