HyukjinKwon commented on code in PR #41770:
URL: https://github.com/apache/spark/pull/41770#discussion_r1259150145
##########
python/pyspark/ml/torch/distributor.py:
##########
@@ -388,35 +391,55 @@ def __init__(
RuntimeError
If an active SparkSession is unavailable.
"""
- super().__init__(num_processes, local_mode, use_gpu)
- self.ssl_conf = "pytorch.spark.distributor.ignoreSsl" # type: ignore
+ super().__init__(num_processes, local_mode, use_gpu,
ssl_conf=_ssl_conf)
self._validate_input_params()
self.input_params = self._create_input_params()
+ @staticmethod
+ def _get_torchrun_args(local_mode: bool, num_processes: int) ->
Tuple[List[Any], int]:
+ """
+ Given the mode and the number of processes, create the arguments to be
given to for torch
+
+ Parameters
+ ---------
+ local_mode: bool
+ Whether or not we are running training locally or in a distributed
fashion
+
+ num_processes: int
+ The number of processes that we are going to use
+
+ Returns
+ ------
+ Tuple[List[Any], int]
Review Comment:
Fix the return type per
https://numpydoc.readthedocs.io/en/latest/format.html#parameters and remove
newlines between parameters
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
