diff options
| author | Marcin Junczys-Dowmunt <marcinjd@microsoft.com> | 2018-11-26 05:34:07 +0300 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2018-11-26 05:34:07 +0300 |
| commit | 3ae12feaaa0ee2b60d1e80e1d159d4799d645640 (patch) | |
| tree | 7038039916ade09b18adff65c292395f526c11af | |
| parent | 44696a0d30be11cca1fd9e2e3207b6b052ecc9cf (diff) | |
Update README.md
| -rw-r--r-- | training-basics-spm/README.md | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/training-basics-spm/README.md b/training-basics-spm/README.md index b96bf82..23f0ca9 100644 --- a/training-basics-spm/README.md +++ b/training-basics-spm/README.md @@ -207,7 +207,7 @@ We are training of four GPUs defined with `--devices 0 1 2 3`. Change this to th ``` The training should stop if cross-entropy on the validation set -stops improving. +stops improving. Depending on the number of and generation of GPUs you are using that may take a while. ### Translating the test and validation sets with evaluation |