Official implementation for LaCo (EMNLP 2024 Findings)

Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)

[ICLR 2022] "Unified Vision Transformer Compression" by Shixing Yu*, Tianlong Chen*, Jiayi Shen, Huan Yuan, Jianchao Tan, Sen Yang, Ji Liu, Zhangyang Wang

Export utility for unconstrained channel pruned models

A method to increase the speed and lower the memory footprint of existing vision transformers.

TinyFusion: Diffusion Transformers Learned Shallow

A family of compressed models obtained via pruning and knowledge distillation

The simplest, fastest repository for training/finetuning medium-sized GPTs.

Official PyTorch implementation of "EdgeSAM: Prompt-In-the-Loop Distillation for On-Device Deployment of SAM"

Official repository for the AAAI2025 paper ( Can We Get Rid of Handcrafted Feature Extractors? SparseViT: Nonsemantics-Centered, Parameter-Efficient Image Manipulation Localization through Spare-Co…

EfficientSAM: Leveraged Masked Image Pretraining for Efficient Segment Anything

[ICML 2024] SqueezeLLM: Dense-and-Sparse Quantization

Code for the ICML 2023 paper "SparseGPT: Massive Language Models Can Be Accurately Pruned in One-Shot".

An official implementation of the paper "How Sparse Can We Prune A Deep Network: A Fundamental Limit Viewpoint".

[NeurIPS 2024] AlphaPruning: Using Heavy-Tailed Self Regularization Theory for Improved Layer-wise Pruning of Large Language Models

Code for the ICLR 2023 paper "GPTQ: Accurate Post-training Quantization of Generative Pretrained Transformers".

The official code of the paper "PyramidDrop: Accelerating Your Large Vision-Language Models via Pyramid Visual Redundancy Reduction".

Pruning the VLLMs

[NeurIPS'23 Oral] Visual Instruction Tuning (LLaVA) built towards GPT-4V level capabilities and beyond.

LLaVA-PruMerge: Adaptive Token Reduction for Efficient Large Multimodal Models

[ECCV 2024 Oral] Code for paper: An Image is Worth 1/2 Tokens After Layer 2: Plug-and-Play Inference Acceleration for Large Vision-Language Models

[NeurIPS 2024] SlimSAM: 0.1% Data Makes Segment Anything Slim

[NeurIPS 2023] LLM-Pruner: On the Structural Pruning of Large Language Models. Support Llama-3/3.1, Llama-2, LLaMA, BLOOM, Vicuna, Baichuan, TinyLlama, etc.

[ICML 2022] "DepthShrinker: A New Compression Paradigm Towards Boosting Real-Hardware Efficiency of Compact Neural Networks", by Yonggan Fu, Haichuan Yang, Jiayi Yuan, Meng Li, Cheng Wan, Raghurama…

Neural Network Compression Framework for enhanced OpenVINO™ inference

A collection of pre-trained, state-of-the-art models in the ONNX format

Official Pytorch Implementation of "Outlier Weighed Layerwise Sparsity (OWL): A Missing Secret Sauce for Pruning LLMs to High Sparsity"

Prospect Pruning: Finding Trainable Weights at Initialization Using Meta-Gradients