hfxunlp/transformer

Neutron: A pytorch based implementation of Transformer and its variants.

PythonGPL-3.0

Readme
3Issues
63Stargazers
8Watchers

Issues

Code of your "Learning Hard Retrieval Decoder Attention for Transformers" paper
#3 opened 2 years ago by HaisongDing
2
more empirical results are expected
#2 opened 4 years ago by alphadl
4
Can't give odd number embedding size for isize
#1 opened 5 years ago by StalVars
1

Topics

neural-machine-translation transformer seq2seq natural-language-processing attention-is-all-you-need average-attention-network dynamic-sentence-sampling robust-neural-machine-translation beam-search ensemble average-models pytorch python3 multi-gpu sentential-context context-aware-nmt relative-position optimizers dynamic-batch-size multilingual-nmt

Contact site admin: Geeks.