AV
A. Van Steenweghen
info
Please Note
<p>This page displays the records of the person named above and is not linked to a unique person identifier. This record may need to be merged to a profile.</p>
2 records found
1
EasyCompress
Automated Compression for Deep Learning Models
Master thesis
(2023)
-
A. Van Steenweghen, L. Miranda da Cruz, Rui Maranhao, A. van Deursen, J.C. van Gemert
Over the past years the size of deep learning models has been growing consistently. This growth has led to significant improvements in performance, but at the expense of increased computational resource demands. Compression techniques can be used to improve the efficiency of deep learning models by shrinking their size and computational needs, while
preserving performance.
This thesis presents EasyCompress, an automated and user-friendly tool to compress deep learning models. The tool improves on existing compression research by focusing on generalizability and practical usability, in three ways. Firstly, it aligns with specific compression objectives and performance requirements, ensuring the compression accomplishes its intended goal effectively. Secondly, it employs flexible compression techniques, so that it is applicable to a diverse set of models without requiring deep model knowledge. Finally, it automates the compression process, eliminating difficult and time-consuming implementation
efforts.
EasyCompress intelligently selects, tailors, and combines various compression techniques to minimize model size, latency, or number of computations while preserving performance. It employs structured pruning to reduce the number of parameters and computations, uses knowledge distillation techniques to ensure better accuracy recovery, and uses quantization to achieve additional compression.
The tool’s effectiveness is evaluated across diverse model architectures and configurations. Experimental results on a range of models and datasets demonstrate its ability to reduce the model size at least 5-fold, inference time by at least 1.5-fold, and the number of computations by at least 3-fold. Most compression rates are even higher, reaching up to 10, 20, and even 100-fold reductions.
The tool is available online at https://thesis.abelvansteenweghen.com. ...
preserving performance.
This thesis presents EasyCompress, an automated and user-friendly tool to compress deep learning models. The tool improves on existing compression research by focusing on generalizability and practical usability, in three ways. Firstly, it aligns with specific compression objectives and performance requirements, ensuring the compression accomplishes its intended goal effectively. Secondly, it employs flexible compression techniques, so that it is applicable to a diverse set of models without requiring deep model knowledge. Finally, it automates the compression process, eliminating difficult and time-consuming implementation
efforts.
EasyCompress intelligently selects, tailors, and combines various compression techniques to minimize model size, latency, or number of computations while preserving performance. It employs structured pruning to reduce the number of parameters and computations, uses knowledge distillation techniques to ensure better accuracy recovery, and uses quantization to achieve additional compression.
The tool’s effectiveness is evaluated across diverse model architectures and configurations. Experimental results on a range of models and datasets demonstrate its ability to reduce the model size at least 5-fold, inference time by at least 1.5-fold, and the number of computations by at least 3-fold. Most compression rates are even higher, reaching up to 10, 20, and even 100-fold reductions.
The tool is available online at https://thesis.abelvansteenweghen.com. ...
Over the past years the size of deep learning models has been growing consistently. This growth has led to significant improvements in performance, but at the expense of increased computational resource demands. Compression techniques can be used to improve the efficiency of deep learning models by shrinking their size and computational needs, while
preserving performance.
This thesis presents EasyCompress, an automated and user-friendly tool to compress deep learning models. The tool improves on existing compression research by focusing on generalizability and practical usability, in three ways. Firstly, it aligns with specific compression objectives and performance requirements, ensuring the compression accomplishes its intended goal effectively. Secondly, it employs flexible compression techniques, so that it is applicable to a diverse set of models without requiring deep model knowledge. Finally, it automates the compression process, eliminating difficult and time-consuming implementation
efforts.
EasyCompress intelligently selects, tailors, and combines various compression techniques to minimize model size, latency, or number of computations while preserving performance. It employs structured pruning to reduce the number of parameters and computations, uses knowledge distillation techniques to ensure better accuracy recovery, and uses quantization to achieve additional compression.
The tool’s effectiveness is evaluated across diverse model architectures and configurations. Experimental results on a range of models and datasets demonstrate its ability to reduce the model size at least 5-fold, inference time by at least 1.5-fold, and the number of computations by at least 3-fold. Most compression rates are even higher, reaching up to 10, 20, and even 100-fold reductions.
The tool is available online at https://thesis.abelvansteenweghen.com.
preserving performance.
This thesis presents EasyCompress, an automated and user-friendly tool to compress deep learning models. The tool improves on existing compression research by focusing on generalizability and practical usability, in three ways. Firstly, it aligns with specific compression objectives and performance requirements, ensuring the compression accomplishes its intended goal effectively. Secondly, it employs flexible compression techniques, so that it is applicable to a diverse set of models without requiring deep model knowledge. Finally, it automates the compression process, eliminating difficult and time-consuming implementation
efforts.
EasyCompress intelligently selects, tailors, and combines various compression techniques to minimize model size, latency, or number of computations while preserving performance. It employs structured pruning to reduce the number of parameters and computations, uses knowledge distillation techniques to ensure better accuracy recovery, and uses quantization to achieve additional compression.
The tool’s effectiveness is evaluated across diverse model architectures and configurations. Experimental results on a range of models and datasets demonstrate its ability to reduce the model size at least 5-fold, inference time by at least 1.5-fold, and the number of computations by at least 3-fold. Most compression rates are even higher, reaching up to 10, 20, and even 100-fold reductions.
The tool is available online at https://thesis.abelvansteenweghen.com.
Sentiment analysis techniques estimate the opinion of the au- thor of a text towards an entity from that text. Current sen- timent analysis techniques are based on language features or deep learning methods. However, they do not make use of the extensive background knowledge that human readers can have. This makes it difficult for these models to detect irony, pop culture references and other subtleties for which connec- tions between entities need to be known. The usage of knowl- edge graphs allows these models to use the enormous exist- ing knowledge bases. We propose a political stance detection pipeline that makes use of knowledge graphs and sentiment analysis. The proposed pipeline uses a combination of exist- ing deep learning methods and classic rule-based methods to train an opinion-aware knowledge graph, with which it clas- sifies sentences as either liberal, conservative or neutral. The pipeline acts as both a classifier and a framework that can in- tegrate existing stance detection models. In an experimental evaluation on the IBC and SemEval datasets, the proposed pipeline achieves an average F-score of 0.63, outperforming traditional machine learning models.
...
Sentiment analysis techniques estimate the opinion of the au- thor of a text towards an entity from that text. Current sen- timent analysis techniques are based on language features or deep learning methods. However, they do not make use of the extensive background knowledge that human readers can have. This makes it difficult for these models to detect irony, pop culture references and other subtleties for which connec- tions between entities need to be known. The usage of knowl- edge graphs allows these models to use the enormous exist- ing knowledge bases. We propose a political stance detection pipeline that makes use of knowledge graphs and sentiment analysis. The proposed pipeline uses a combination of exist- ing deep learning methods and classic rule-based methods to train an opinion-aware knowledge graph, with which it clas- sifies sentences as either liberal, conservative or neutral. The pipeline acts as both a classifier and a framework that can in- tegrate existing stance detection models. In an experimental evaluation on the IBC and SemEval datasets, the proposed pipeline achieves an average F-score of 0.63, outperforming traditional machine learning models.