A Multi-Teraflop Constituency Parser using GPUs
By John F. Canny, David Leo Wright Hall, Daniel Klein. 2013
22 citations recorded by the source corpus.
Constituency parsing with rich grammars remains a computational challenge. Graphics Processing Units (GPUs) have previously been used to accelerate CKY chart evaluation, but gains over CPU parsers were modest. In this paper, we describe a collection of new techniques that enable chart evaluation at close to the GPU’s practical maximum speed (a Teraflop), or around a half-trillion rule evaluations per second. Net parser performance on a 4-GPU system is over 1 thousand length30 sentences/second (1 trillion rules/sec), and 400 general sentences/second for the Berkeley Parser Grammar. The techniques we introduce include grammar compilation, recursive symbol blocking, and cache-sharing.
Subfields: Artificial Intelligence, Natural Language Processing
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Luke S. Zettlemoyer
- Michael Lewis
- Mohit Bansal
- Percy Liang
- Graham Neubig
- Yejin Choi
- Jianfeng Gao
- Noah A. Smith
In Natural Language Processing
- Luke S. Zettlemoyer
- Michael Lewis
- Mohit Bansal
- Graham Neubig
- Percy Liang
- Yejin Choi
- Daniel Klein
- Noah A. Smith
Citations in the corpus
- Billion-Scale Similarity Search with GPUs (2017, methodology): “The CPU analogy is that there are up to 255 vector registers of width 32, with warp lanes as SIMD vector lanes.”
- LSTM CCG Parsing (2016, methodology): “Other work has successfully applied GPUs to parsing, but has required GPU-specific code and algorithms (Yi et al., 2011; Johnson, 2011; Canny et al., 2013; Hall et al., 2014).”
- Sparser, Better, Faster GPU Parsing (2014, methodology): “Apart from the model of Canny et al. (2013), there have been a few attempts at using GPUs in NLP contexts before.”