Views
No views yet
glm-10b is pretrained on the Pile dataset. It has 48 transformer layers, with hidden size 4096 and 64 attention heads in each layer. The model is pretrained with autoregressive blank filling objectives designed for natural language understanding, seq2seq, and language modeling. Find more details from our repo.[MASK] for short blank filling, [sMASK] for sentence filling, and [gMASK] for left to right generation. You can find examples about different masks from here. The prediction always begin with a special <|startofpiece|> token and ends with a <|endofpiece|> token.@article{DBLP:conf/acl/DuQLDQY022,
author = {Zhengxiao Du and
Yujie Qian and
Xiao Liu and
Ming Ding and
Jiezhong Qiu and
Zhilin Yang and
Jie Tang},
title = {{GLM:} General Language Model Pretraining with Autoregressive Blank Infilling},
booktitle = {Proceedings of the 60th Annual Meeting of the Association for Computational
Linguistics (Volume 1: Long Papers), {ACL} 2022, Dublin, Ireland,
May 22-27, 2022},
pages = {320--335},
publisher = {Association for Computational Linguistics},
year = {2022},
}