p85971406
/
mindspore
forked from mindspore-Ecosystem/mindspore
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Reset datasink to false when training with ps using ascend

This commit is contained in:
ZPaC 2023-02-09 10:53:32 +08:00
parent 5c5ca060ba
commit b81da9d274
1 changed files with 9 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
mindspore/python/mindspore/train/model.py
View File

@ -1009,6 +1009,11 @@ class Model:
... loss_scale_manager=loss_scale_manager)
>>> model.train(2, dataset)
"""
device_target = context.get_context("device_target")
if _is_ps_mode() and not _cache_enable() and (device_target in ["Ascend", "CPU"]) and dataset_sink_mode:
logger.info("For PS mode, reset datasink mode to False when using Ascend or CPU backend.")
dataset_sink_mode = False
Validator.check_bool(dataset_sink_mode)
if isinstance(self._train_network, nn.GraphCell) and dataset_sink_mode:
raise ValueError("Dataset sink mode is currently not supported when training with a GraphCell.")
@ -1144,6 +1149,10 @@ class Model:
>>> model = Model(net, loss_fn=loss, optimizer=optim, metrics={"accuracy"})
>>> model.fit(2, train_dataset, valid_dataset)
"""
device_target = context.get_context("device_target")
if _is_ps_mode() and not _cache_enable() and (device_target in ["Ascend", "CPU"]) and dataset_sink_mode:
logger.info("For PS mode, reset datasink mode to False when using Ascend or CPU backend.")
dataset_sink_mode = False
dataset_sink_mode = Validator.check_bool(dataset_sink_mode)
valid_dataset_sink_mode = Validator.check_bool(valid_dataset_sink_mode)