Picture for Tomas Hrycej

Tomas Hrycej

Efficient Neural Network Training via Subset Pretraining

Add code
Oct 21, 2024
Viaarxiv icon

Reducing the Transformer Architecture to a Minimum

Add code
Oct 17, 2024
Viaarxiv icon

Make Deep Networks Shallow Again

Add code
Sep 15, 2023
Viaarxiv icon

Number of Attention Heads vs Number of Transformer-Encoders in Computer Vision

Add code
Sep 15, 2022
Figure 1 for Number of Attention Heads vs Number of Transformer-Encoders in Computer Vision
Figure 2 for Number of Attention Heads vs Number of Transformer-Encoders in Computer Vision
Figure 3 for Number of Attention Heads vs Number of Transformer-Encoders in Computer Vision
Figure 4 for Number of Attention Heads vs Number of Transformer-Encoders in Computer Vision
Viaarxiv icon

Training Neural Networks in Single vs Double Precision

Add code
Sep 15, 2022
Figure 1 for Training Neural Networks in Single vs Double Precision
Figure 2 for Training Neural Networks in Single vs Double Precision
Figure 3 for Training Neural Networks in Single vs Double Precision
Viaarxiv icon

Representational Capacity of Deep Neural Networks -- A Computing Study

Add code
Jul 19, 2019
Figure 1 for Representational Capacity of Deep Neural Networks -- A Computing Study
Figure 2 for Representational Capacity of Deep Neural Networks -- A Computing Study
Figure 3 for Representational Capacity of Deep Neural Networks -- A Computing Study
Viaarxiv icon

Singular Value Decomposition and Neural Networks

Add code
Jun 27, 2019
Figure 1 for Singular Value Decomposition and Neural Networks
Figure 2 for Singular Value Decomposition and Neural Networks
Figure 3 for Singular Value Decomposition and Neural Networks
Figure 4 for Singular Value Decomposition and Neural Networks
Viaarxiv icon