no code implementations • 16 Apr 2024 • Haozheng Fan, Hao Zhou, Guangtai Huang, Parameswaran Raman, Xinwei Fu, Gaurav Gupta, Dhananjay Ram, Yida Wang, Jun Huan
In this paper, we showcase HLAT: a 7 billion parameter decoder-only LLM pre-trained using trn1 instances over 1. 8 trillion tokens.
no code implementations • 19 Oct 2023 • Qingru Zhang, Dhananjay Ram, Cole Hawkins, Sheng Zha, Tuo Zhao
These models leverage the attention mechanism to capture long- and short-range dependencies in the sequence.
no code implementations • 19 Nov 2019 • Dhananjay Ram, Lesly Miculicich, Hervé Bourlard
Here, we show that the CNN based matching outperforms DTW based matching using bottleneck features as well.
no code implementations • 30 Jun 2019 • Dhananjay Ram, Lesly Miculicich, Hervé Bourlard
State of the art solutions to query by example spoken term detection (QbE-STD) usually rely on bottleneck feature representation of the query and audio document to perform dynamic time warping (DTW) based template matching.
2 code implementations • EMNLP 2018 • Lesly Miculicich, Dhananjay Ram, Nikolaos Pappas, James Henderson
Neural Machine Translation (NMT) can be improved by including document-level contextual information.
1 code implementation • NAACL 2018 • Lesly Miculicich Werlen, Nikolaos Pappas, Dhananjay Ram, Andrei Popescu-Belis
Neural sequence-to-sequence networks with attention have achieved remarkable performance for machine translation.
no code implementations • 19 Oct 2016 • Dhananjay Ram, Debasis Kundu, Rajesh M. Hegde
In this work, a Bayesian approach to speaker normalization is proposed to compensate for the degradation in performance of a speaker independent speech recognition system.