Search Results for author:
Found 7 papers, 0 papers with code
Can the capability of Large Language Models be described by human ability? A Meta Study
The evaluation benchmarks are categorized into 6 primary abilities and 11 sub-abilities in human aspect.
A General Error-Theoretical Analysis Framework for Constructing Compression Strategies
The exponential growth in parameter size and computational complexity of deep models poses significant challenges for efficient deployment.
Compression for Better: A General and Stable Lossless Compression Framework
A key challenge is effectively leveraging compression errors and defining the boundaries for lossless compression to minimize model loss.
Lossless Model Compression via Joint Low-Rank Factorization Optimization
Low-rank factorization is a popular model compression technique that minimizes the error $\delta$ between approximated and original weight matrices.
FP=xINT:A Low-Bit Series Expansion Algorithm for Post-Training Quantization
We introduce a deep model series expansion framework to address this issue, enabling rapid and accurate approximation of unquantized models without calibration sets or fine-tuning.
Mixed-Precision Inference Quantization: Radically Towards Faster inference speed, Lower Storage requirement, and Lower Loss
Based on the model's resilience to computational noise, model quantization is important for compressing models and improving computing speed.
Quantization in Layer's Input is Matter
In this paper, we will show that the quantization in layer's input is more important than parameters' quantization for loss function.
