Zobrazeno 1 - 10
of 1 938
pro vyhledávání: '"Gusak, A."'
Autor:
Abakumov, Serhii, Gusak, Andriy
Formation of the intermediate phase patterns in the thin-film co-deposition process is simulated using the Stochastic Kinetic Mean-Field method and Monte Carlo. Three basic morphologies of the 2D sections are distinguished: (1) spots (rod-like in 3D)
Externí odkaz:
http://arxiv.org/abs/2412.03457
Autor:
Gusak, Andriy, Abakumov, Serhii
The Ising model is well-known for illustrating the fundamental characteristics of phase transitions in closed systems. In this article, we propose a generalization of the two-dimensional Ising model to open systems, considering the divergence of exte
Externí odkaz:
http://arxiv.org/abs/2411.00584
Scalability issue plays a crucial role in productionizing modern recommender systems. Even lightweight architectures may suffer from high computational overload due to intermediate calculations, limiting their practicality in real-world applications.
Externí odkaz:
http://arxiv.org/abs/2409.18721
The goal of modern sequential recommender systems is often formulated in terms of next-item prediction. In this paper, we explore the applicability of generative transformer-based models for the Top-K sequential recommendation task, where the goal is
Externí odkaz:
http://arxiv.org/abs/2409.17730
Scalability is a major challenge in modern recommender systems. In sequential recommendations, full Cross-Entropy (CE) loss achieves state-of-the-art recommendation quality but consumes excessive GPU memory with large item catalogs, limiting its prac
Externí odkaz:
http://arxiv.org/abs/2408.02354
Autor:
Cherniuk, Daria, Abukhovich, Stanislav, Phan, Anh-Huy, Oseledets, Ivan, Cichocki, Andrzej, Gusak, Julia
Tensor decomposition of convolutional and fully-connected layers is an effective way to reduce parameters and FLOP in neural networks. Due to memory and power consumption limitations of mobile or embedded devices, the quantization step is usually nec
Externí odkaz:
http://arxiv.org/abs/2308.04595
We propose Rockmate to control the memory requirements when training PyTorch DNN models. Rockmate is an automatic tool that starts from the model code and generates an equivalent model, using a predefined amount of memory for activations, at the cost
Externí odkaz:
http://arxiv.org/abs/2307.01236
Large-scale transformer models have shown remarkable performance in language modelling tasks. However, such models feature billions of parameters, leading to difficulties in their deployment and prohibitive training costs from scratch. To reduce the
Externí odkaz:
http://arxiv.org/abs/2306.02697
Autor:
Ya. V. Krylova, E. V. Kondakova, A. N. Gavrilenko, A. M. Chekalov, L. V. Fedorova, L. V. Stelmakh, E. V. Babenko, T. S. Shchegoleva, A. A. Gusak, V. V. Baykov, N. B. Mikhailova, A. D. Kulagin
Publikováno v:
Онкогематология, Vol 19, Iss 3, Pp 153-158 (2024)
Recent advances in the diagnosis and understanding of follicular lymphoma (FL) pathogenesis have had a significant impact on therapeutic tactics. The life expectancy of patients has increased significantly. Currently, the 5-year overall survival of F
Externí odkaz:
https://doaj.org/article/561967c1236a4dbea2b3ce4004bf3c32
