A hybrid architecture with two QRNN layers followed by a PAR Transformer reaches 1.013 BPC on enwik8 and 20.91 PPL on Wikitext-103 with 41-60M parameters.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Efficient Language Modeling for Low-Resource Settings with Hybrid RNN-Transformer Architectures
A hybrid architecture with two QRNN layers followed by a PAR Transformer reaches 1.013 BPC on enwik8 and 20.91 PPL on Wikitext-103 with 41-60M parameters.