https://github.com/KwangHoonAn/Quantizations
Quantizing deep convolutional networks for efficient inference: A whitepaper
태스크의 하위 분류와 논문·코드·평가·기법·데이터셋을 탐색하세요.
Quantization is a promising technique to reduce the computation cost of neural network training, which can replace high-cost floating-point numbers (e.g., float32) with low-cost fixed-point numbers (e.g., int8/int16).
Source: Adaptive Precision Training: Quantify Back Propagation in Neural Networks with Fixed-point Numbers
관련 논문과 평가 결과에 연결된 구현입니다. 출처에서 확인된 저자 구현에는 공식 코드 표시가 있습니다.
Quantizing deep convolutional networks for efficient inference: A whitepaper
VideoBERT: A Joint Model for Video and Language Representation Learning
HAQ: Hardware-Aware Automated Quantization with Mixed Precision
An empirical study of LLaMA3 quantization: from LLMs to MLLMs
Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference
Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference
Randomized Quantization: A Generic Augmentation for Data Agnostic Self-supervised Learning
FALCON: Feature-Label Constrained Graph Net Collapse for Memory Efficient GNNs
Unsupervised Cross-lingual Representation Learning for Speech Recognition
Randomized Quantization: A Generic Augmentation for Data Agnostic Self-supervised Learning
TinySAM: Pushing the Envelope for Efficient Segment Anything Model
SignSGD with Federated Defense: Harnessing Adversarial Attacks through Gradient Sign Decoding
Fate: Fast Edge Inference of Mixture-of-Experts Models via Cross-Layer Gate
SignSGD with Federated Defense: Harnessing Adversarial Attacks through Gradient Sign Decoding
Randomized Quantization is All You Need for Differential Privacy in Federated Learning
Incremental Network Quantization: Towards Lossless CNNs with Low-Precision Weights
출처: Papers with Code · CC BY-SA 4.0. 아카이브 자료와 새로 수집한 논문은 별도로 관리합니다.