'--user-dir' documentation (correct) (#447)

Summary:
Command line option --user-dir documented in docs/overview.rst
Pull Request resolved: https://github.com/pytorch/fairseq/pull/447

Differential Revision: D13674744

Pulled By: myleott

fbshipit-source-id: 17049ee5c9f692f5298ef9fa7381ee583f269cde
This commit is contained in:
Davide Caroselli 2019-01-15 11:16:26 -08:00 committed by Facebook Github Bot
parent b15f5f5384
commit ebaf8c5030

View File

@ -46,3 +46,25 @@ decorators, for example::
Once registered, new plug-ins can be used with the existing :ref:`Command-line
Tools`. See the Tutorial sections for more detailed walkthroughs of how to add
new plug-ins.
**Loading plug-ins from another directory**
New plug-ins can be defined in a custom module stored in the user system. In order to import the module, and make the plugin available to *fairseq*, the command line supports the ``--user-dir`` flag that can be used to specify a custom location for additional modules to load into *fairseq*.
For example, assuming this directory tree::
/home/user/my-module/
└── __init__.py
with ``__init__.py``::
from fairseq.models import register_model_architecture
from fairseq.models.transformer import transformer_vaswani_wmt_en_de_big
@register_model_architecture('transformer', 'my_transformer')
def transformer_mmt_big(args):
transformer_vaswani_wmt_en_de_big(args)
it is possible to invoke the ``train.py`` script with the new architecture with::
python3 train.py ... --user-dir /home/user/my-module -a my_transformer --task translation