G
George E. Dahl
Researcher at Google
Publications - 66
Citations - 36393
George E. Dahl is an academic researcher from Google. The author has contributed to research in topics: Artificial neural network & Hidden Markov model. The author has an hindex of 36, co-authored 56 publications receiving 29759 citations. Previous affiliations of George E. Dahl include Microsoft & University of Toronto.
Papers
More filters
Posted Content
Faster Neural Network Training with Data Echoing
TL;DR: This paper introduces "data echoing," which reduces the total computation used by earlier pipeline stages and speeds up training whenever computation upstream from accelerators dominates the training time.
Proceedings ArticleDOI
The Importance of Generation Order in Language Modeling
TL;DR: This article studied the influence of token generation order on model quality via a two-pass language model that produces partially-filled sentence templates and then fills in missing tokens, and found that the most effective strategy generates function words in the first pass followed by content words.
Dissertation
Deep learning approaches to problems in speech recognition, computational chemistry, and natural language text processing
TL;DR: A new neural network generative model of parsed sentences capable of generating reasonable samples is introduced and demonstrated, which results in a model for molecular activity prediction substantially more effective than production systems used in the pharmaceutical industry.
Posted Content
Parallel Architecture and Hyperparameter Search via Successive Halving and Classification.
TL;DR: This work presents a simple and powerful algorithm for parallel black box optimization called Successive Halving and Classification (SHAC), which operates in stages of parallel function evaluations and trains a cascade of binary classifiers to iteratively cull the undesirable regions of the search space.
Posted Content
The Importance of Generation Order in Language Modeling
TL;DR: The authors studied the influence of token generation order on model quality via a two-pass language model that produces partially-filled sentence templates and then fills in missing tokens, and found that the most effective strategy generates function words in the first pass followed by content words.