Zobrazeno 1 - 10
of 68
pro vyhledávání: '"Hannun, Awni"'
Large language models have emerged as a versatile tool but are challenging to apply to tasks lacking large inference budgets and large in-domain training sets. This work formalizes these constraints and distinguishes four important variables: the pre
Externí odkaz:
http://arxiv.org/abs/2402.01093
Large neural networks pretrained on web-scale corpora are central to modern machine learning. In this paradigm, the distribution of the large, heterogeneous pretraining data rarely matches that of the application domain. This work considers modifying
Externí odkaz:
http://arxiv.org/abs/2311.11973
Large, pre-trained models are problematic to use in resource constrained applications. Fortunately, task-aware structured pruning methods offer a solution. These approaches reduce model size by dropping structural units like layers and attention head
Externí odkaz:
http://arxiv.org/abs/2311.06382
Autor:
Kahn, Jacob, Pratap, Vineel, Likhomanenko, Tatiana, Xu, Qiantong, Hannun, Awni, Cai, Jeff, Tomasello, Paden, Lee, Ann, Grave, Edouard, Avidov, Gilad, Steiner, Benoit, Liptchinsky, Vitaliy, Synnaeve, Gabriel, Collobert, Ronan
As the computational requirements for machine learning systems and the size and complexity of machine learning frameworks increases, essential framework innovation has become challenging. While computational needs have driven recent compiler, network
Externí odkaz:
http://arxiv.org/abs/2201.12465
We develop an algorithm which can learn from partially labeled and unsegmented sequential data. Most sequential loss functions, such as Connectionist Temporal Classification (CTC), break down when many labels are missing. We address this problem with
Externí odkaz:
http://arxiv.org/abs/2201.12208
Finite-state transducers (FSTs) are frequently used in speech recognition. Transducer composition is an essential operation for combining different sources of information at different granularities. However, composition is also one of the more comput
Externí odkaz:
http://arxiv.org/abs/2110.02848
Autor:
Knott, Brian, Venkataraman, Shobha, Hannun, Awni, Sengupta, Shubho, Ibrahim, Mark, van der Maaten, Laurens
Secure multi-party computation (MPC) allows parties to perform computations on data while keeping that data private. This capability has great potential for machine-learning applications: it facilitates training of machine-learning models on private
Externí odkaz:
http://arxiv.org/abs/2109.00984
Autor:
Hannun, Awni
The decade from 2010 to 2020 saw remarkable improvements in automatic speech recognition. Many people now use speech recognition on a daily basis, for example to perform voice search queries, send text messages, and interact with voice assistants lik
Externí odkaz:
http://arxiv.org/abs/2108.00084
Autor:
Hannun, Awni
Machine intelligence can develop either directly from experience or by inheriting experience through evolution. The bulk of current research efforts focus on algorithms which learn directly from experience. I argue that the alternative, evolution, is
Externí odkaz:
http://arxiv.org/abs/2106.11151
Autor:
Shi, Yuge, Seely, Jeffrey, Torr, Philip H. S., Siddharth, N., Hannun, Awni, Usunier, Nicolas, Synnaeve, Gabriel
Machine learning systems typically assume that the distributions of training and test sets match closely. However, a critical requirement of such systems in the real world is their ability to generalize to unseen domains. Here, we propose an inter-do
Externí odkaz:
http://arxiv.org/abs/2104.09937