p47068915
/
mindspore
forked from mindspore-Ecosystem/mindspore
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

!27059 Fix document API 

Merge pull request !27059 from huangxinjing/add_api_doc
This commit is contained in:
i-robot 2021-12-02 10:51:15 +00:00 committed by Gitee
parent 18d9eb8a5e 3fe30ac2a0
commit d2f3d6a793
2 changed files with 25 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
mindspore/communication/__init__.py
View File

@ -13,7 +13,13 @@
# limitations under the License. # limitations under the License.
# ============================================================================ # ============================================================================
""" """
Collective communication interface. Collective communication interface. Note the API in the file needs to preset communication environment variables. For
the Ascend cards, users need to prepare the rank table, set rank_id and device_id. Please see the `Ascend tutorial \
<https://www.mindspore.cn/tutorials/zh-CN/master/intermediate/distributed_training/
distributed_training_ascend.html#id12>`_ for more details.
For the GPU device, users need to prepare the host file and mpi, please see the `GPU tutorial \
<https://www.mindspore.cn/tutorials/zh-CN/r1.5/intermediate/distributed_training/distributed_training_gpu.html#id6>`_
for more details.
""" """
from .management import GlobalComm, init, release, get_rank, get_group_size, get_world_rank_from_group_rank, \ from .management import GlobalComm, init, release, get_rank, get_group_size, get_world_rank_from_group_rank, \

28
mindspore/communication/management.py
View File

@ -89,7 +89,8 @@ def init(backend_name=None):
Note: Note:
The full name of HCCL is Huawei Collective Communication Library. The full name of HCCL is Huawei Collective Communication Library.
The full name of NCCL is NVIDIA Collective Communication Library. The full name of NCCL is NVIDIA Collective Communication Library.
This method should be used after set_context. This method should be used after set_context. The user needs to preset communication environment variables
before running the following example, please see the docstring of the mindspore.managerment.
Args: Args:
backend_name (str): Backend, using HCCL/NCCL. If the `backend_name` is None, system will recognize backend_name (str): Backend, using HCCL/NCCL. If the `backend_name` is None, system will recognize
@ -153,7 +154,8 @@ def release():
Release distributed resource. e.g. HCCL/NCCL. Release distributed resource. e.g. HCCL/NCCL.
Note: Note:
This method should be used after init(). This method should be used after init(). The user needs to preset communication environment variables
before running the following example, please see the docstring of the mindspore.managerment.
Raises: Raises:
RuntimeError: If failed to release distributed resource. RuntimeError: If failed to release distributed resource.
@ -171,7 +173,8 @@ def get_rank(group=GlobalComm.WORLD_COMM_GROUP):
Get the rank ID for the current device in the specified collective communication group. Get the rank ID for the current device in the specified collective communication group.
Note: Note:
This method should be used after init(). This method should be used after init(). The user needs to preset communication environment variables
before running the following example, please see the docstring of the mindspore.managerment.
Args: Args:
group (str): The communication group to work on. Normally, the group should be created by create_group, group (str): The communication group to work on. Normally, the group should be created by create_group,
@ -203,7 +206,8 @@ def get_local_rank(group=GlobalComm.WORLD_COMM_GROUP):
Note: Note:
GPU version of MindSpore doesn't support this method. GPU version of MindSpore doesn't support this method.
This method should be used after init(). This method should be used after init(). The user needs to preset communication environment variables
before running the following example, please see the docstring of the mindspore.managerment.
Args: Args:
group (str): The communication group to work on. Normally, the group should be created by create_group, group (str): The communication group to work on. Normally, the group should be created by create_group,
@ -237,7 +241,8 @@ def get_group_size(group=GlobalComm.WORLD_COMM_GROUP):
Get the rank size of the specified collective communication group. Get the rank size of the specified collective communication group.
Note: Note:
This method should be used after init(). This method should be used after init(). The user needs to preset communication environment variables before
running the following example, please see the docstring of the mindspore.managerment.
Args: Args:
group (str): The communication group to work on. Normally, the group should be created by create_group, group (str): The communication group to work on. Normally, the group should be created by create_group,
@ -272,7 +277,8 @@ def get_local_rank_size(group=GlobalComm.WORLD_COMM_GROUP):
Note: Note:
GPU version of MindSpore doesn't support this method. GPU version of MindSpore doesn't support this method.
This method should be used after init(). This method should be used after init(). The user needs to preset communication environment variables before
running the following example, please see the docstring of the mindspore.managerment.
Args: Args:
group (str): The communication group to work on. The group is created by create_group group (str): The communication group to work on. The group is created by create_group
@ -308,7 +314,8 @@ def get_world_rank_from_group_rank(group, group_rank_id):
Note: Note:
GPU version of MindSpore doesn't support this method. GPU version of MindSpore doesn't support this method.
The parameter group should not be "hccl_world_group". The parameter group should not be "hccl_world_group".
This method should be used after init(). This method should be used after init(). The user needs to preset communication environment variables
before running the following example, please see the docstring of the mindspore.managerment.
Args: Args:
group (str): The communication group to work on. The group is created by create_group. group (str): The communication group to work on. The group is created by create_group.
@ -349,6 +356,8 @@ def get_group_rank_from_world_rank(world_rank_id, group):
GPU version of MindSpore doesn't support this method. GPU version of MindSpore doesn't support this method.
The parameter group should not be "hccl_world_group". The parameter group should not be "hccl_world_group".
This method should be used after init(). This method should be used after init().
The user needs to preset communication environment variables before running the following example, please see
the docstring of the mindspore.managerment.
Args: Args:
world_rank_id (int): A rank ID in the world communication group. world_rank_id (int): A rank ID in the world communication group.
@ -386,12 +395,11 @@ def create_group(group, rank_ids):
Note: Note:
GPU version of MindSpore doesn't support this method. GPU version of MindSpore doesn't support this method.
The size of rank_ids should be larger than 1, rank_ids should not have duplicate data. The size of rank_ids should be larger than 1, rank_ids should not have duplicate data.
This method should be used after init(). This method should be used after init().
Only support global single communication group in PyNative mode. Only support global single communication group in PyNative mode.
The user needs to preset communication environment variables before running the following example, please see
the docstring of the mindspore.managerment.
Args: Args:
group (str): The name of the communication group to be created. group (str): The name of the communication group to be created.