Trellis Networks for Sequence Modeling
This work addresses the problem of improving sequence modeling performance for tasks such as language modeling, offering a novel hybrid approach that could benefit researchers and practitioners in natural language processing.
The authors introduced trellis networks, a new architecture for sequence modeling that combines elements from temporal convolutional and recurrent networks, and demonstrated that it outperforms state-of-the-art methods on benchmarks like word-level and character-level language modeling tasks, including stress tests for long-term memory retention.
We present trellis networks, a new architecture for sequence modeling. On the one hand, a trellis network is a temporal convolutional network with special structure, characterized by weight tying across depth and direct injection of the input into deep layers. On the other hand, we show that truncated recurrent networks are equivalent to trellis networks with special sparsity structure in their weight matrices. Thus trellis networks with general weight matrices generalize truncated recurrent networks. We leverage these connections to design high-performing trellis networks that absorb structural and algorithmic elements from both recurrent and convolutional models. Experiments demonstrate that trellis networks outperform the current state of the art methods on a variety of challenging benchmarks, including word-level language modeling and character-level language modeling tasks, and stress tests designed to evaluate long-term memory retention. The code is available at https://github.com/locuslab/trellisnet .