Fix pretrained models table

This commit is contained in:
Lysandre
2019-11-26 15:47:02 -05:00
parent cf26a0c85e
commit ce02550d50

View File

@@ -190,7 +190,7 @@ Here is the full list of the currently provided pretrained models together with
| | ``albert-xxlarge-v2`` | | 12 repeating layer, 128 embedding, 4096-hidden, 64-heads, 223M parameters |
| | | | ALBERT xxlarge model with no dropout, additional training data and longer training |
| | | (see `details <https://github.com/google-research/google-research/tree/master/albert>`__) |
+--------------------------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
+-------------------+------------------------------------------------------------+---------------------------------------------------------------------------------------------------------------------------------------+
.. <https://huggingface.co/transformers/examples.html>`__