HyperAI超神经

首页资讯最新论文教程数据集百科 SOTA LLM 模型天梯 GPU 天梯顶会

中文

HyperAI超神经

Language Modelling On Hutter Prize

评估指标

Bit per Character (BPC)

Number of params

评测结果

各个模型在此基准测试上的表现结果

模型名称	Bit per Character (BPC)	Number of params	Paper Title	Repository
Large RHN	1.27	46M	Recurrent Highway Networks	-
Large FS-LSTM-4	1.245	47M	Fast-Slow Recurrent Neural Networks	-
Transformer-XL + RMS dynamic eval	0.94	277M	Dynamic Evaluation of Transformer Language Models	-
18-layer Transformer-XL	1.03	88M	Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context	-
Large mLSTM +emb +WN +VD	1.24	46M	Multiplicative LSTM for sequence modelling	-
Mogrifier LSTM	1.122	96M	Mogrifier LSTM	-
12-layer Transformer-XL	1.06	41M	Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context	-
64-layer Character Transformer Model	1.06	235M	Character-Level Language Modeling with Deeper Self-Attention	-
3-layer AWD-LSTM	1.232	47M	An Analysis of Neural Language Modeling at Multiple Scales	-
Longformer Small	1.00	41M	Longformer: The Long-Document Transformer	-
12-layer Character Transformer Model	1.11	44M	Character-Level Language Modeling with Deeper Self-Attention	-
FS-LSTM-4	1.277	27M	Fast-Slow Recurrent Neural Networks	-
mLSTM + dynamic eval	1.08	46M	Dynamic Evaluation of Neural Sequence Models	-
Longformer Large	0.99	102M	Longformer: The Long-Document Transformer	-
RHN - depth 5 [zilly2016recurrent]	1.31	-	Recurrent Highway Networks	-
24-layer Transformer-XL	0.99	277M	Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context	-
Compressive Transformer	0.97	-	Compressive Transformers for Long-Range Sequence Modelling	-
Mogrifier LSTM + dynamic eval	0.988	96M	Mogrifier LSTM	-

0 of 18 row(s) selected.