Behavior-gated language models, which use a fixed behavior classifier's outputs to gate word predictions, achieve small perplexity reductions on therapy dialogues and Penn Treebank.
Partially Shuffling the Training Data to Improve Language Models
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
Although SGD requires shuffling the training data between epochs, currently none of the word-level language modeling systems do this. Naively shuffling all sentences in the training data would not permit the model to learn inter-sentence dependencies. Here we present a method that partially shuffles the training data between epochs. This method makes each batch random, while keeping most sentence ordering intact. It achieves new state of the art results on word-level language modeling on both the Penn Treebank and WikiText-2 datasets.
fields
cs.CL 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Behavior Gated Language Models
Behavior-gated language models, which use a fixed behavior classifier's outputs to gate word predictions, achieve small perplexity reductions on therapy dialogues and Penn Treebank.