riaz.somc/transformers
1
0
Fork 0
mirror of https://github.com/huggingface/transformers.git synced 2025-08-01 18:51:14 +06:00
Code Issues Actions 3 Packages Projects Releases Wiki Activity

[testing doc] bring doc up to date (#11359)

Browse Source 
* bring doc up to date

* fix
This commit is contained in:
Stas Bekman 2021-04-21 08:51:00 -07:00 committed by GitHub
parent 41f3133a3a
commit 9f72e8f4e1
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 8 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

18
docs/source/testing.rst
View File

@ -502,20 +502,18 @@ Distributed training
thing and end up thinking they are ``pytest`` and start running the test suite in loops. It works, however, if one thing and end up thinking they are ``pytest`` and start running the test suite in loops. It works, however, if one
spawns a normal process that then spawns off multiple workers and manages the IO pipes. spawns a normal process that then spawns off multiple workers and manages the IO pipes.
This is still under development but you can study 2 different tests that perform this successfully: Here are some tests that use it:
* :prefix_link:`test_seq2seq_examples_multi_gpu.py <examples/seq2seq/test_seq2seq_examples_multi_gpu.py>` - a * :prefix_link:`test_trainer_distributed.py <tests/test_trainer_distributed.py>`
``pytorch-lightning``-running test (had to use PL's ``ddp`` spawning method which is the default) * :prefix_link:`test_deepspeed.py <tests/deepspeed/test_deepspeed.py>`
* :prefix_link:`test_finetune_trainer.py <examples/seq2seq/test_finetune_trainer.py>` - a normal (non-PL) test
To jump right into the execution point, search for the ``execute_subprocess_async`` function in those tests. To jump right into the execution point, search for the ``execute_subprocess_async`` call in those tests.
You will need at least 2 GPUs to see these tests in action: You will need at least 2 GPUs to see these tests in action:
.. code-block:: bash .. code-block:: bash
CUDA_VISIBLE_DEVICES="0,1" RUN_SLOW=1 pytest -sv examples/seq2seq/test_finetune_trainer.py \ CUDA_VISIBLE_DEVICES=0,1 RUN_SLOW=1 pytest -sv tests/test_trainer_distributed.py
examples/seq2seq/test_seq2seq_examples_multi_gpu.py
Output capture Output capture
@ -718,10 +716,10 @@ To start using those all you need is to make sure that the test resides in a sub
from transformers.testing_utils import TestCasePlus from transformers.testing_utils import TestCasePlus
class PathExampleTest(TestCasePlus): class PathExampleTest(TestCasePlus):
def test_something_involving_local_locations(self): def test_something_involving_local_locations(self):
data_dir = self.examples_dir / "seq2seq/test_data/wmt_en_ro" data_dir = self.tests_dir / "fixtures/tests_samples/wmt_en_ro"
If you don't need to manipulated paths via ``pathlib`` or you just need a path as a string, you can always invoked If you don't need to manipulate paths via ``pathlib`` or you just need a path as a string, you can always invoked
``str()`` on the ``pathlib`` oboject or use the accessors ending with ``_str``. For example: ``str()`` on the ``pathlib`` object or use the accessors ending with ``_str``. For example:
.. code-block:: python .. code-block:: python