Zobrazeno 1 - 10
of 18
pro vyhledávání: '"Ramaswamy, Swaroop"'
Autor:
Breiner, Theresa, Ramaswamy, Swaroop, Variani, Ehsan, Garg, Shefali, Mathews, Rajiv, Sim, Khe Chai, Gupta, Kilol, Chen, Mingqing, McConnaughey, Lara
Personalization of speech models on mobile devices (on-device personalization) is an active area of research, but more often than not, mobile devices have more text-only data than paired audio-text data. We explore training a personalized language mo
Externí odkaz:
http://arxiv.org/abs/2207.00706
Autor:
Amid, Ehsan, Ganesh, Arun, Mathews, Rajiv, Ramaswamy, Swaroop, Song, Shuang, Steinke, Thomas, Suriyakumar, Vinith M., Thakkar, Om, Thakurta, Abhradeep
In this paper, we revisit the problem of using in-distribution public data to improve the privacy/utility trade-offs for differentially private (DP) model training. (Here, public data refers to auxiliary data sets that have no privacy concerns.) We d
Externí odkaz:
http://arxiv.org/abs/2112.00193
Autor:
Dang, Trung, Thakkar, Om, Ramaswamy, Swaroop, Mathews, Rajiv, Chin, Peter, Beaufays, Françoise
Distributed learning paradigms such as federated learning often involve transmission of model updates, or gradients, over a network, thereby avoiding transmission of private data. However, it is possible for sensitive information about the training d
Externí odkaz:
http://arxiv.org/abs/2111.00556
Autor:
Dang, Trung, Thakkar, Om, Ramaswamy, Swaroop, Mathews, Rajiv, Chin, Peter, Beaufays, Françoise
End-to-end Automatic Speech Recognition (ASR) models are commonly trained over spoken utterances using optimization methods like Stochastic Gradient Descent (SGD). In distributed settings like Federated Learning, model training requires transmission
Externí odkaz:
http://arxiv.org/abs/2104.07815
Autor:
Ramaswamy, Swaroop, Thakkar, Om, Mathews, Rajiv, Andrew, Galen, McMahan, H. Brendan, Beaufays, Françoise
This paper presents the first consumer-scale next-word prediction (NWP) model trained with Federated Learning (FL) while leveraging the Differentially Private Federated Averaging (DP-FedAvg) technique. There has been prior work on building practical
Externí odkaz:
http://arxiv.org/abs/2009.10031
Recent works have shown that generative sequence models (e.g., language models) have a tendency to memorize rare or unique sequences in the training data. Since useful models are often trained on sensitive data, to ensure the privacy of the training
Externí odkaz:
http://arxiv.org/abs/2006.07490
Autor:
Augenstein, Sean, McMahan, H. Brendan, Ramage, Daniel, Ramaswamy, Swaroop, Kairouz, Peter, Chen, Mingqing, Mathews, Rajiv, Arcas, Blaise Aguera y
To improve real-world applications of machine learning, experienced modelers develop intuition about their datasets, their models, and how the two interact. Manual inspection of raw data - of representative samples, of outliers, of misclassifications
Externí odkaz:
http://arxiv.org/abs/1911.06679
We show that a word-level recurrent neural network can predict emoji from text typed on a mobile keyboard. We demonstrate the usefulness of transfer learning for predicting emoji by pretraining the model using a language modeling task. We also propos
Externí odkaz:
http://arxiv.org/abs/1906.04329
Existing approaches for training neural networks with user-level differential privacy (e.g., DP Federated Averaging) in federated learning (FL) settings involve bounding the contribution of each user's model update by clipping it to some constant val
Externí odkaz:
http://arxiv.org/abs/1905.03871
Autor:
Hard, Andrew, Rao, Kanishka, Mathews, Rajiv, Ramaswamy, Swaroop, Beaufays, Françoise, Augenstein, Sean, Eichner, Hubert, Kiddon, Chloé, Ramage, Daniel
We train a recurrent neural network language model using a distributed, on-device learning framework called federated learning for the purpose of next-word prediction in a virtual keyboard for smartphones. Server-based training using stochastic gradi
Externí odkaz:
http://arxiv.org/abs/1811.03604