A case for application-based linear transformers : applied towards text-to-spectrogram

Graduate Thesis Or Dissertation

Público Deposited

Citeable URL: https://ir.library.oregonstate.edu/concern/graduate_thesis_or_dissertations/3x816v73n

Descriptions

Attribute Name	Values
Creator	Agostinelli, Victor
Abstract	Various natural language processing (NLP) tasks necessitate deep models that are fast, efficient, and small based on their ultimate application at the edge or elsewhere. While significant investigation has furthered the efficiency and reduced the size of these models, reducing their downstream latency without significant trade-offs remains a difficult task. This thesis proposes the modular utilization of state-of-the-art attention linearization techniques that results in fully linearized run-time with respect to the size of the sample at inference for autoregressive decoding tasks for neural machine translation (NMT), text-to-spectrogram (TTS) tasks, and other applications while minimizing the associated costs to downstream model performance.
License	All rights reserved
Resource Type	Masters Thesis
Fecha de Emisión	2023-04-14
Degree Level	Master's
Degree Name	Master of Science (M.S.)
Degree Field	Electrical and Computer Engineering
Degree Grantor	Oregon State University
Commencement Year	2023
Advisor	Chen, Lizhong
Committee Member	Abbas, Houssam Lee, Ben Tejasvi, Anand
Academic Affiliation	Electrical Engineering and Computer Science
Declaración de derechos	In Copyright
Publisher	Oregon State University
Peer Reviewed	No
Language	English [eng]

Parents:

This work has no parents.

En Collection:

Miniatura	Título	Fecha de subida	Visibilidad	Acciones
	VictorAgostinelli2023.pdf	2023-06-06	Público	Descargar