OSSForks/pytorch
1
0
Fork 0
mirror of https://github.com/zebrajr/pytorch.git synced 2025-12-08 07:39:33 +01:00
Code Issues Actions 38 Packages Projects Releases Wiki Activity

[Docs] Fix docstring format (#99396)

Browse Source 
Fixes #ISSUE_NUMBER

Pull Request resolved: https://github.com/pytorch/pytorch/pull/99396
Approved by: https://github.com/awgu
This commit is contained in:
zhouzaida 2023-04-28 01:10:03 +00:00 committed by PyTorch MergeBot
parent 64efd88845
commit b51f92ebda
2 changed files with 11 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
torch/distributed/fsdp/fully_sharded_data_parallel.py
View File

@ -216,7 +216,7 @@ class FullyShardedDataParallel(nn.Module, _FSDPState):
Args:
module (nn.Module):
This is the module to be wrapped with FSDP.
process_group: Optional[Union[ProcessGroup, Tuple[ProcessGroup, ProcessGroup]]]
process_group (Optional[Union[ProcessGroup, Tuple[ProcessGroup, ProcessGroup]]]):
This is the process group used for collective communications and
the one over which the model is sharded. For hybrid sharding strategies such as
``ShardingStrategy.HYBRID_SHARD`` users can
@ -1458,7 +1458,7 @@ class FullyShardedDataParallel(nn.Module, _FSDPState):
corresponding to the unflattened parameters and holding the
sharded optimizer state.
model (torch.nn.Module):
Refer to :meth:``shard_full_optim_state_dict``.
Refer to :meth:`shard_full_optim_state_dict`.
optim (torch.optim.Optimizer): Optimizer for ``model`` 's
parameters.
@ -1785,7 +1785,7 @@ class FullyShardedDataParallel(nn.Module, _FSDPState):
) -> Dict[str, Any]:
"""
This hook is intended be used by ``torch.distributed.NamedOptimizer``.
The functionality is identical to ``:meth:optim_state_dict`` except
The functionality is identical to :meth:`optim_state_dict` except
for the different arguments.
Args:
@ -1916,7 +1916,7 @@ class FullyShardedDataParallel(nn.Module, _FSDPState):
) -> Dict[str, Any]:
"""
This hook is intended be used by ``torch.distributed.NamedOptimizer``.
The functionality is identical to ``:meth:optim_state_dict_to_load``
The functionality is identical to :meth:`optim_state_dict_to_load`
except for the different arguments.
Args: