From 4c5e592a2cec4e00b6d6d8f437525be1661a13de Mon Sep 17 00:00:00 2001 From: Yada Pruksachatkun Date: Sun, 3 Jun 2018 17:24:24 -0400 Subject: [PATCH] Added some debugging notes for future coders I spent quite a bit of time debugging why .score()'s output did not line up with my personal observations with the model performance so wanted to put some more documentation on this here. --- README.rst | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/README.rst b/README.rst index c14925d..7b6041f 100644 --- a/README.rst +++ b/README.rst @@ -58,6 +58,11 @@ For more information, see the `documentation `_. +Debugging Notes +--------------- + +If you're using the .score() method, note that it might not be evaluating/scoring the right granularity (for example, evaluating letter by letter versus word by word when your model outputs words). Also check out whole_sequence_accuracy() method. + |Travis|_ .. |Travis| image:: https://api.travis-ci.org/larsmans/seqlearn.png?branch=master