List your company

Language Modeling on WikiText-2 v1 (val)

42.41PerplexityAWD-LSTM-MoS + dynamical evaluation
34.749286.4596138.17189.8804May 14, 2019Jul 2, 2020Aug 21, 2021Oct 11, 2022Nov 30, 2023Jan 18, 2025Mar 10, 2026

Evaluation results

MethodLinks
AWD-LSTM-MoS + dynamical evaluation#Param=35M, dynamic evaluation=true2019.0542.41—
AWD-LSTM-DRILL + dynamic evaluation#Param=34M, dynamic evaluation=true2019.0543.9—
AWD-LSTM + dynamic evaluation#Param=33M, dynamic evaluation=true2019.0546.4—
AWD-LSTM + continuous cache pointer#Param=33M, dynamic evaluation=true2019.0553.8—
Ensemble of AllWeight=12023.1155.4—
AWD-LSTM-DOCWeight=0.412023.1160.27—
AWD-LSTM-MoS#Param=35M2019.0563.88—
AWD-LSTM-MOSWeight=0.172023.1163.92—
AWD-LSTM-DRILL#Param=34M, finetuned=true2019.0564.9—
AWD-LSTM-DRILL#Param=34M, finetuned=false2019.0565.7—
Transformer-XLWeight=0.162023.1167.41—
AWD-LSTM#Param=33M, finetuned=true2019.0568.6—
2-layer skip connection LSTM#Param=24M2019.0569.1—
AWD-LSTM#Param=33M, finetuned=false2019.0569.1—
QRNNWeight=0.002023.1169.23—
EGRUWeight=0.222023.1169.4—
Fast WeightsWeight=0.032023.1169.51—
Variational LSTM + augmented loss#Param=28M2019.0591.5—
Fuzzy-Gated + RPAcontext=768, evaluation protocol=sequential, stride=no-overlap (stride=context)2026.03189.85.246
KnerserNey-5gramWeight=0.012023.11233.93—