Search Results for author:
Found 37 papers, 22 papers with code
PTSBench: A Comprehensive Post-Training Sparsity Benchmark Towards Algorithms and Models
Our PTSBench can provide (1) new observations for a better understanding of the PTS algorithms, (2) in-depth and comprehensive evaluations for the sparsification ability of models, and (3) a well-structured and easy-integrate open-source framework.
HarmoniCa: Harmonizing Training and Inference for Better Feature Caching in Diffusion Transformer Acceleration
Diffusion Transformers (DiTs) excel in generative tasks but face practical deployment challenges due to high inference costs.
A Survey of Low-bit Large Language Models: Basics, Systems, and Algorithms
Large language models (LLMs) have achieved remarkable advancements in natural language processing, showcasing exceptional performance across various tasks.
OmniBal: Towards Fast Instruct-tuning for Vision-Language Models via Omniverse Computation Balance
We rebalanced the computational loads from data, model, and memory perspectives to address this issue, achieving more balanced computation across devices.
Temporal Feature Matters: A Framework for Diffusion Model Quantization
However, unlike traditional models, diffusion models critically rely on the time-step for the multi-round denoising.
ME-Switch: A Memory-Efficient Expert Switching Framework for Large Language Models
LLM development involves pre-training a foundation model on massive data, followed by fine-tuning on task-specific data to create specialized experts.
Selective Focus: Investigating Semantics Sensitivity in Post-training Quantization for Lane Detection
In this paper, we pioneeringly investigate semantic sensitivity to post-processing for lane detection with a novel Lane Distortion Score.
LLMC: Benchmarking Large Language Model Quantization with a Versatile Compression Toolkit
In this paper, we present LLMC, a plug-and-play compression toolkit, to fairly and systematically explore the impact of quantization.
Fast and Controllable Post-training Sparsity: Learning Optimal Sparsity Allocation with Global Constraint in Minutes
Previous methods for finding sparsity rates mainly focus on the training-aware scenario, which usually fails to converge stably under the PTS setting with limited data and much less training cost.
2023 Low-Power Computer Vision Challenge (LPCVC) Summary
This article describes the 2023 IEEE Low-Power Computer Vision Challenge (LPCVC).
ProPD: Dynamic Token Tree Pruning and Generation for LLM Parallel Decoding
Additionally, it introduces a dynamic token tree generation algorithm to balance the computation and parallelism of the verification phase in real-time and maximize the overall efficiency across different batch sizes, sequence lengths, and tasks, etc.
TFMQ-DM: Temporal Feature Maintenance Quantization for Diffusion Models
Remarkably, our quantization approach, for the first time, achieves model performance nearly on par with the full-precision model under 4-bit weight quantization.
QLLM: Accurate and Efficient Low-Bitwidth Quantization for Large Language Models
Additionally, an adaptive strategy is designed to autonomously determine the optimal number of sub-channels for channel disassembly.
Lossy and Lossless (L$^2$) Post-training Model Size Compression
Then, a dedicated differentiable counter is introduced to guide the optimization of lossy compression to arrive at a more suitable point for later lossless compression.
SysNoise: Exploring and Benchmarking Training-Deployment System Inconsistency
Extensive studies have shown that deep learning models are vulnerable to adversarial and natural noises, yet little is known about model robustness on noises caused by different system implementations.
Outlier Suppression+: Accurate quantization of large language models by equivalent and optimal shifting and scaling
The channel-wise shifting aligns the center of each channel for removal of outlier asymmetry.
Lossy and Lossless (L2) Post-training Model Size Compression
Then, a dedicated differentiable counter is introduced to guide the optimization of lossy compression to arrive at a more suitable point for later lossless compression.
Exploring the Relationship Between Architectural Design and Adversarially Robust Generalization
In particular, we comprehensively evaluated 20 most representative adversarially trained architectures on ImageNette and CIFAR-10 datasets towards multiple l_p-norm adversarial attacks.
Annealing-Based Label-Transfer Learning for Open World Object Detection
To the best of our knowledge, this is the first OWOD work without manual unknown selection.
Exploring the Relationship between Architecture and Adversarially Robust Generalization
Inparticular, we comprehensively evaluated 20 most representative adversarially trained architectures on ImageNette and CIFAR-10 datasets towards multiple `p-norm adversarial attacks.
Outlier Suppression: Pushing the Limit of Low-bit Transformer Language Models
With the trends of large NLP models, the increasing memory and computation costs hinder their efficient deployment on resource-limited devices.
QDrop: Randomly Dropping Quantization for Extremely Low-bit Post-Training Quantization
With QDROP, the limit of PTQ is pushed to the 2-bit activation for the first time and the accuracy boost can be up to 51. 49%.
MQBench: Towards Reproducible and Deployable Model Quantization Benchmark
Surprisingly, no existing algorithm wins every challenge in MQBench, and we hope this work could inspire future research directions.
Distribution-sensitive Information Retention for Accurate Binary Neural Network
We present a novel Distribution-sensitive Information Retention Network (DIR-Net) that retains the information in the forward and backward propagation by improving internal propagation and introducing external representations.
RobustART: Benchmarking Robustness on Architecture Design and Training Techniques
Thus, we propose RobustART, the first comprehensive Robustness investigation benchmark on ImageNet regarding ARchitecture design (49 human-designed off-the-shelf architectures and 1200+ networks from neural architecture search) and Training techniques (10+ techniques, e. g., data augmentation) towards diverse noises (adversarial, natural, and system noises).
Real World Robustness from Systematic Noise
Systematic error, which is not determined by chance, often refers to the inaccuracy (involving either the observation or measurement process) inherent to a system.
A Free Lunch From ANN: Towards Efficient, Accurate Spiking Neural Networks Calibration
Moreover, our calibration algorithm can produce SNN with state-of-the-art architecture on the large-scale ImageNet dataset, including MobileNet and RegNet.
Diversifying Sample Generation for Accurate Data-Free Quantization
Unfortunately, we find that in practice, the synthetic data identically constrained by BN statistics suffers serious homogenization at both distribution level and sample level and further causes a significant performance drop of the quantized model.
BRECQ: Pushing the Limit of Post-Training Quantization by Block Reconstruction
To further employ the power of quantization, the mixed precision technique is incorporated in our framework by approximating the inter-layer and intra-layer sensitivity.
MixMix: All You Need for Data-Free Compression Are Feature and Data Mixing
However, the inversion process only utilizes biased feature statistics stored in one model and is from low-dimension to high-dimension.
Once Quantization-Aware Training: High Performance Extremely Low-bit Architecture Search
Therefore, we propose to combine Network Architecture Search methods with quantization to enjoy the merits of the two sides.
Binary Neural Networks: A Survey
The binary neural network, largely saving the storage and computation, serves as a promising technique for deploying deep models on resource-limited devices.
Efficient Bitwidth Search for Practical Mixed Precision Neural Network
Network quantization has rapidly become one of the most widely used methods to compress and accelerate deep neural networks.
Towards Unified INT8 Training for Convolutional Neural Network
In this paper, we give an attempt to build a unified 8-bit (INT8) training framework for common convolutional neural networks from the aspects of both accuracy and speed.
Balanced Binary Neural Networks with Gated Residual
In this paper, we attempt to maintain the information propagated in the forward process and propose a Balanced Binary Neural Networks with Gated Residual (BBG for short).
Ranked #1049 on
Image Classification
on ImageNet
(Top 1 Accuracy metric)
Forward and Backward Information Retention for Accurate Binary Neural Networks
Our empirical study indicates that the quantization brings information loss in both forward and backward propagation, which is the bottleneck of training accurate binary neural networks.
Differentiable Soft Quantization: Bridging Full-Precision and Low-Bit Neural Networks
Hardware-friendly network quantization (e. g., binary/uniform quantization) can efficiently accelerate the inference and meanwhile reduce memory consumption of the deep neural networks, which is crucial for model deployment on resource-limited devices like mobile phones.
