SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

minor typo fix

Browse Source 
*negative* log-likelihood
This commit is contained in:
Joe Davison
2021-04-01 11:58:37 -06:00
committed by GitHub
parent 57c1749efa
commit f4ad3d8cea
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
docs/source/perplexity.rst
View File

@@ -18,8 +18,8 @@ that the metric applies specifically to classical language models (sometimes cal
models) and is not well defined for masked language models like BERT (see :doc:`summary of the models models) and is not well defined for masked language models like BERT (see :doc:`summary of the models
<model_summary>`). <model_summary>`).
Perplexity is defined as the exponentiated average log-likelihood of a sequence. If we have a tokenized sequence Perplexity is defined as the exponentiated average negative log-likelihood of a sequence. If we have a tokenized
:math:`X = (x_0, x_1, \dots, x_t)`, then the perplexity of :math:`X` is, sequence :math:`X = (x_0, x_1, \dots, x_t)`, then the perplexity of :math:`X` is,
.. math:: .. math::