https://github.com/wang-h/neuralcompressor
Compressing Word Embeddings via Deep Compositional Code Learning
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.
Compressing Word Embeddings via Deep Compositional Code Learning
ActNN: Reducing Training Memory Footprint via 2-Bit Activation Compressed Training
wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations
Generating Diverse Structure for Image Inpainting With Hierarchical VQ-VAE
Unsupervised Person Re-identification via Softened Similarity Learning
LogicNets: Co-Designed Neural Networks and Circuits for Extreme-Throughput Applications
SpikeLLM: Scaling up Spiking Neural Network to Large Language Models via Saliency-based Spiking
Hyper-Sphere Quantization: Communication-Efficient SGD for Federated Learning
YOLOv6: A Single-Stage Object Detection Framework for Industrial Applications
Training High-Performance and Large-Scale Deep Neural Networks with Full 8-bit Integers
SpeechT5: Unified-Modal Encoder-Decoder Pre-Training for Spoken Language Processing
LSQ+: Improving low-bit quantization through learnable offsets and better initialization
RMP-Loss: Regularizing Membrane Potential Distribution for Spiking Neural Networks
RMP-Loss: Regularizing Membrane Potential Distribution for Spiking Neural Networks
FP6-LLM: Efficiently Serving Large Language Models Through FP6-Centric Algorithm-System Co-Design
TOD: GPU-accelerated Outlier Detection via Tensor Operations
출처: Papers with Code · CC BY-SA 4.0. 아카이브 자료와 새로 수집한 논문은 별도로 관리합니다.