diff options
| author | Yiding Wang <wydwww@outlook.com> | 2016-09-13 10:52:03 +0300 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2016-09-13 10:52:03 +0300 |
| commit | 5abe83bdfa0efa14f9ce5e75880f5269115127c2 (patch) | |
| tree | 5aea0c1b1ed7b8d92bc118229ff53e8926cbbed4 | |
| parent | 9d7b9ea4c9ba38e92dc0186af33ff7c0f323d2a4 (diff) | |
Just typos
| -rw-r--r-- | doc/criterion.md | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/doc/criterion.md b/doc/criterion.md index 270edb9..babd39c 100644 --- a/doc/criterion.md +++ b/doc/criterion.md @@ -96,7 +96,7 @@ criterion.sizeAverage = false criterion = nn.ClassNLLCriterion([weights]) ``` -The negative log likelihood criterion. It is useful to train a classication problem with `n` classes. +The negative log likelihood criterion. It is useful to train a classification problem with `n` classes. If provided, the optional argument `weights` should be a 1D `Tensor` assigning weight to each of the classes. This is particularly useful when you have an unbalanced training set. @@ -143,7 +143,7 @@ criterion = nn.CrossEntropyCriterion([weights]) This criterion combines [`LogSoftMax`](#nn.LogSoftMax) and [`ClassNLLCriterion`](#nn.ClassNLLCriterion) in one single class. -It is useful to train a classication problem with `n` classes. +It is useful to train a classification problem with `n` classes. If provided, the optional argument `weights` should be a 1D `Tensor` assigning weight to each of the classes. This is particularly useful when you have an unbalanced training set. The `input` given through a `forward()` is expected to contain scores for each class: `input` has to be a 1D `Tensor` of size `n`. |