[bugfix] Fix signature mismatch in DDPCPUHPCAccelerator's model_to_device (#5505)

* Update ddp_cpu_hpc_accelerator.py

* Update CHANGELOG.md
This commit is contained in:
ananthsub 2021-01-16 02:57:30 -08:00 committed by Jirka Borovec
parent 293984bbc0
commit ad4e25fde5
2 changed files with 2 additions and 2 deletions

View File

@ -208,6 +208,7 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
- Logging only on `not should_accumulate()` during training ([#5417](https://github.com/PyTorchLightning/pytorch-lightning/pull/5417))
- Resolve interpolation bug with Hydra ([#5406](https://github.com/PyTorchLightning/pytorch-lightning/pull/5406))
- Check environ before selecting a seed to prevent warning message ([#4743](https://github.com/PyTorchLightning/pytorch-lightning/pull/4743))
- Fixed signature mismatch in `model_to_device` of `DDPCPUHPCAccelerator` ([#5505](https://github.com/PyTorchLightning/pytorch-lightning/pull/5505))
## [1.1.3] - 2021-01-05

View File

@ -36,8 +36,7 @@ class DDPCPUHPCAccelerator(DDPHPCAccelerator):
super().__init__(trainer, cluster_environment, ddp_plugin)
self.nickname = 'ddp_cpu'
def model_to_device(self, model, process_idx):
# Todo: required argument `process_idx` is not used
def model_to_device(self, model):
model.cpu()
def get_device_ids(self):