Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / guolinke/tupe issues and pull requests
#22 - Definition of two learnable parameters
Issue -
State: open - Opened by tom68-ll over 2 years ago
- 4 comments
#21 - 我终于完成数据预处理了,谢谢您的回答和帮助,我想问一下按照你的参数在4块16g的v100上要跑多久能跑完?
Issue -
State: open - Opened by tom68-ll over 2 years ago
- 4 comments
#20 - Improve pos_embed calculation
Pull Request -
State: open - Opened by ZhiyuanChen over 3 years ago
#19 - How to set up input for sentence classification task?
Issue -
State: closed - Opened by huberemanuel over 3 years ago
- 2 comments
#18 - No such file or directory: './data-bin/pt-data-0508/dict.txt'
Issue -
State: open - Opened by huberemanuel over 3 years ago
- 4 comments
#17 - question of providing bert model integrated with TUPE
Issue -
State: open - Opened by EchoFei333 over 3 years ago
- 1 comment
#16 - The first column of `dict.txt` is equivalent to the vocabulary learned from fastBPE?
Issue -
State: closed - Opened by huberemanuel over 3 years ago
- 2 comments
#15 - How to calculate correlation in Figure 2?
Issue -
State: open - Opened by Redaimao over 3 years ago
- 1 comment
#14 - Fix bug of warnings not imported in mha
Pull Request -
State: closed - Opened by ZhiyuanChen over 3 years ago
#13 - Fix bug of pad not defined in multihead_attention
Pull Request -
State: closed - Opened by ZhiyuanChen over 3 years ago
#12 - Discrepancy between the paper and the implementation?
Issue -
State: closed - Opened by tonyswoo over 3 years ago
- 2 comments
#11 - Reproduce BERT and TUPE
Issue -
State: closed - Opened by wyu97 almost 4 years ago
- 3 comments
#10 - About the pre-trained checkpoints
Issue -
State: open - Opened by VerdureChen almost 4 years ago
- 1 comment
#9 - some issues
Issue -
State: open - Opened by sanwei111 almost 4 years ago
- 1 comment
#8 - Do these codes also fit for encoder-decoder transformer?
Issue -
State: open - Opened by SefaZeng about 4 years ago
- 3 comments
#7 - 有没验证过,是pretrain阶段的增益,还是本身这个position-encoding就有增益(不pretrain也有)?
Issue -
State: closed - Opened by guotong1988 over 4 years ago
- 4 comments
#6 - Could you please point out the CORE code for TUPE for study? As there are too many fairseq code.
Issue -
State: closed - Opened by guotong1988 over 4 years ago
- 1 comment
#5 - A TensorFlow Implementation?
Issue -
State: closed - Opened by guotong1988 over 4 years ago
- 1 comment
#4 - About the reset function in attention bias
Issue -
State: closed - Opened by Erutan-pku over 4 years ago
- 4 comments
#3 - What's the format of the raw data?
Issue -
State: open - Opened by Howal over 4 years ago
- 4 comments
#2 - 我跑你的代码,数据集那些怎么弄,怎么装载字典,我没有dict.txt
Issue -
State: open - Opened by wymxz over 4 years ago
- 12 comments
#1 - 相对位置具体配置
Issue -
State: closed - Opened by PROoshio over 4 years ago
- 1 comment