google-research / google-research/google-research
How to obtain representation from Smith, for some arbitrary doc
Open
- Dominant language
- Jupyter Notebook
- Stars
- 38.8k
- Forks
- 8.5k
- PR merge metrics
- No merged PRs in 30d
Description
Hi,
I have looked at the code, but it is not exactly clear on how to obtain representation. Is this something to obtain from layers.get_masked_lm_output, after the NN transformation?
In the paper you suggest that you take the first block representation, but it is not exactly clear how you extract the CLS token from the code at the document level
Many thanks
Contributor guide
Assessment
This issue has not been assessed yet.