added state_dict/load_state_dict for ReduceLROnPlateau #7201
Conversation
kahne
requested review from
apaszke,
colesbury,
ezyang,
gchanan,
soumith and
zdevito
as code owners
|
Thanks @kahne. Is there a test we can add for this? |
|
@ezyang I tested it manually with |
| self.is_better = partial(self._cmp, mode, threshold_mode, threshold) | ||
|
|
||
| def state_dict(self): | ||
| return {key: value for key, value in self.__dict__.items() if key not in {'optimizer', 'is_better'}} |
This comment was marked as off-topic.
This comment was marked as off-topic.
Sorry, something went wrong.
This comment was marked as off-topic.
This comment was marked as off-topic.
Sorry, something went wrong.
This comment was marked as off-topic.
This comment was marked as off-topic.
Sorry, something went wrong.
|
@kahne there are a few tests in |
|
@pytorchbot retest this please |
|
Thanks @kahne! |
|
Hi! I'm trying to use state_dict() on ReduceLROnPlateau but I got this error: I'm using pytorch 0.4.0. Thanks! |
|
@ignaciorlando Hi! This feature is available in 0.4.1 but not 0.4.0. Can you upgrade to 0.4.1 and try again? |
4 participants
There is a merged PR #5300 trying to make ReduceLROnPlateau serializable and resolve issue #5218. However, it's still not working. Moreover, it's more consistent to provide the common state_dict/load_state_dict interface for serialization.