Search Results for author:
Found 8 papers, 0 papers with code
SkipViT: Speeding Up Vision Transformers with a Token-Level Skip Connection
Our method does not add any parameters to the ViT model and aims to find the best trade-off between training throughput and achieving a 0% loss in the Top-1 accuracy of the final model.
SwiftLearn: A Data-Efficient Training Method of Deep Learning Models using Importance Sampling
In this paper, we present SwiftLearn, a data-efficient approach to accelerate training of deep learning models using a subset of data samples selected during the warm-up stages of training.
GQKVA: Efficient Pre-training of Transformers by Grouping Queries, Keys, and Values
We tested our method on ViT, which achieved an approximate 0. 3% increase in accuracy while reducing the model size by about 4% in the task of image classification.
Speeding up Resnet Architecture with Layers Targeted Low Rank Decomposition
Compression of a neural network can help in speeding up both the training and the inference of the network.
Training Acceleration of Low-Rank Decomposed Networks using Sequential Freezing and Rank Quantization
Low Rank Decomposition (LRD) is a model compression technique applied to the weight tensors of deep learning models in order to reduce the number of trainable parameters and computational complexity.
Improving Resnet-9 Generalization Trained on Small Datasets
This paper presents our proposed approach that won the first prize at the ICLR competition on Hardware Aware Efficient Training.
Ensemble-based Adaptive Single-shot Multi-box Detector
This uses data to reduce the uncertainty in the selection of best aspect ratios for the default boxes and improves performance of SSD for datasets containing small and complex objects (e. g., equipments at construction sites).
Efficient Single-Shot Multibox Detector for Construction Site Monitoring
Asset monitoring in construction sites is an intricate, manually intensive task, that can highly benefit from automated solutions engineered using deep neural networks.
