Our new X account is live! Follow @wizwand_team for updates

Language Modeling on Penn Treebank (val)

46.64Perplexity

GL-LWGC-AWD-MOS-LSTM + dynamic evaluation

Updated 3d ago

Evaluation Results

Method	Links
GL-LWGC-AWD-MOS-LSTM + dynamic evaluation 2017.08		46.64
AWD-LSTM-MoS+PDR + dynamic evaluation 2018.08		48
AWD-LSTM-MoS + dynamic evaluation 2018.08		48.3
AWD-LSTM-MoS + dynamic evaluation 2017.08		48.33
AWD-LSTM-MoS 2017.11		48.33
AWD-LSTM-MoS + dynamic evaluation 2019.05		48.33
Ensemble of All 2023.11		48.92
AWD-LSTM-DRILL + dynamic evaluation 2019.05		49.5
AWD-LSTM+PDR + dynamic evaluation 2018.08		50.1
AWD-LSTM + dynamic evaluation 2017.08		51.6
AWD-LSTM + dynamic eval 2017.09		51.6
AWD-LSTM 2017.11		51.6
AWD-LSTM + dynamic evaluation 2018.08		51.6
AWD-LSTM + dynamic evaluation 2019.05		51.6
AWD-LSTM+PDR + continuous cache pointer 2018.08		52.4
DOC + Finetune* + Partial Shuffle 2019.03		53.79
AWD-LSTM - 3-layer LSTM (tied) + continuous cache pointer 2017.08		53.9
AWD-LSTM + neural cache 2017.09		53.9
AWD-LSTM 2017.11		53.9
AWD-LSTM + continuous cache pointer 2018.08		53.9
AWD-LSTM + continuous cache pointer 2019.05		53.9
AWD-LSTM-DOC 2023.11		54.12
DOC + Finetune* 2019.03		54.18
GL-LWGC-AWD-MOS-LSTM + finetune 2017.08		54.24
DOC + Finetune + Partial Shuffle 2019.03		54.3
GL-LWGC-AWD-MOS-LSTM 2017.08		54.57
DOC + Finetune 2019.03		54.62
DOC + Partial Shuffle 2019.03		54.9
2-layers GL-LWGC-AWD-MOS-LSTM + finetune 2017.08		55.18
DOC 2019.03		55.39
MoS + Finetune + Partial Shuffle 2019.03		55.89
AWD-LSTM-MOS 2023.11		56.04
AWD-LSTM-MoS+PDR 2018.08		56.2
AWD-LSTM-MoS 2018.08		56.5
AWD-LSTM-MoS + finetune 2017.08		56.54
AWD-LSTM-MoS 2017.11		56.54
AWD-LSTM-MoS 2019.05		56.54
MoS + Finetune 2019.03		56.76
Reference Model 2017.08		57.41
MoS + Partial Shuffle 2019.03		57.43
AWD-LSTM+PDR 2018.08		57.9
Transformer-XL 2023.11		57.93
AWD-LSTM-MoS 2017.08		58.08
AWD-LSTM-MoS 2017.11		58.08
AWD-LSTM-DRILL 2019.05		58.2
MoS 2019.03		58.27
Fast Weights 2023.11		58.49
Fraternal dropout + AWD-LSTM 3-layer 2017.10		58.9
AWD-LSTM-DRILL 2019.05		59.6
AWD-LSTM 2017.09		59.8
AWD-LSTM 2017.08		60
AWD-LSTM - 3-layer LSTM (tied) 2017.08		60
AWD-LSTM 2017.09		60
AWD-LSTM 2017.11		60
Merity et al. (2017a) - AWD-LSTM 3-layer (baseline) 2017.10		60
AWD-LSTM 2018.08		60
AWD-LSTM 2019.05		60
Lookahead (ADAM) 2019.07		60.28
QRNN 2023.11		60.38
AWD-LSTM 2017.11		60.7
AWD-LSTM 2019.05		60.7
2-layer skip connection LSTM 2017.08		60.9
4-layer skip connection LSTM (tied) 2017.08		60.9
LSTM + BB tuning 2017.09		60.9
2-layer skip connection LSTM 2017.11		60.9
Melis et al. (2017) - 4-layer skip connection LSTM 2017.10		60.9
2-layer skip connection LSTM 2019.05		60.9
Polyak 2019.07		61.18
EGRU 2023.11		61.21
ADAM 2019.07		61.64
Melis et al. (2017) - 5-layer RHN 2017.10		64.8
Lookahead (SGD) 2019.07		65.1
SGD 2019.07		66
Variational RHN + WT 2017.08		67.9
Variational RHN + WT 2016.07		67.9
Variational RHN (tied) 2017.08		67.9
Variational RHN 2017.09		67.9
Variational RHN 2017.11		67.9
Zilly et al. (2016) - Variational RHN 2017.10		67.9
Variational RHN 2019.05		67.9
VD-RHN +RE 2016.11		68.1
Inan et al. (2016) - Variational RHN 2017.10		68.1
VD-LSTM +REAL (large) 2016.11		71.1
Variational LSTM + WT + augmented loss 2016.07		71.1
Variational LSTM (tied) + augmented loss 2017.08		71.1
Variational LSTM + augmented loss 2017.09		71.1
Inan et al. (2016) - Variational LSTM 2017.10		71.1
VD-RHN 2016.11		71.2
Variational RHN 2016.07		71.2
Large regularized LSTM 2014.09		71.9
38 Large LSTMs 2016.11		71.9
Pointer Sentinel-LSTM(medium) 2016.11		72.4
Pointer Sentinel-LSTM 2016.07		72.4
Pointer Sentinel-LSTM 2017.08		72.4
Pointer sentinel-LSTM 2017.09		72.4
Pointer Sentinel-LSTM 2017.11		72.4
Pointer Sentinel-LSTM 2019.05		72.4
Large regularized LSTM 2014.09		72.8
LSTM + dynamic eval 2017.09		73.5
LSTM + dynamic eval 2017.09		74.3

Showing 100 of 178 rows