efficient-model 开源项目推荐（44 个）

mit-han-lab /temporal-shift-module

[ICCV TSM：用于高效视频理解的时间偏移模块

mit-han-lab /once-for-all

[ICLR 2020] Once for All: Train One Network and Specialize it for Efficient Deployment

mit-han-lab /proxylessnas

[ICLR 2019] ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware

mit-han-lab /amc

[ECCV 2018] AMC: AutoML for Model Compression and Acceleration on Mobile Devices

SqueezeAILab /KVQuant

[NeurIPS 2024] KVQuant: Towards 10 Million Context Length LLM Inference with KV Cache Quantization

mit-han-lab /haq

[CVPR 2019, Oral] HAQ: Hardware-Aware Automated Quantization with Mixed Precision

microsoft /nn-Meter

A DNN inference latency prediction toolkit for accurately modeling and predicting the latency on diverse edge devices.

mit-han-lab /hardware-aware-transformers

[ACL'20] HAT：面向高效自然语言处理的硬件感知 Transformer

amirgholami /ZeroQ

[CVPR'20] ZeroQ：一种新颖的零样本量化框架

kssteven418 /I-BERT

[ICML'21 口头报告] I-BERT：仅整数的 BERT 量化

mit-han-lab /amc-models

[ECCV 2018] AMC：面向移动设备模型压缩与加速的 AutoML

youngwanLEE /VoV3D

用于时序建模的高效 3D 主干网络

d-li14 /HBONet

[ICCV 2019] 两个正交维度上的和谐瓶颈，超越 MobileNetV2

kssteven418 /LTP

[KDD'22] 面向 Transformer 的学习式词元剪枝

xvyaward /owq

AAAI 2024 口头报告论文《OWQ：面向大语言模型高效微调与推理的离群值感知权重量化》的代码。

szq0214 /S2-BNN

S2-BNN：通过引导分布校准弥合自监督实数网络与 1-bit 神经网络之间的差距（CVPR 2021）

SHI-Labs /Any-Precision-DNNs

任意精度深度神经网络（AAAI 2021）

HVision-NKU /OffSeg

[ICCV 2025] 重新审视高效语义分割：学习偏移以实现更好的空间与类别特征对齐

lironui /ABCNet

遥感图像的语义分割

mit-han-lab /neurips-micronet

[JMLR'20] NeurIPS 2019 MicroNet 挑战赛高效语言建模冠军