github
/
ColossalAI
mirror of https://github.com/hpcaitech/ColossalAI
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

[fix] fixed the collective pattern name for consistency (#1649) 

* [fix] fixed the collective pattern name for consistency

* polish code
pull/1650/head
Frank Lee 2022-09-26 16:39:37 +08:00 committed by GitHub
parent b2b2a4af98
commit 154d3ef432
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
2 changed files with 7 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
colossalai/tensor/shape_consistency.py
View File

@ -235,7 +235,7 @@ class CollectiveCommPattern(Enum):
GATHER_FWD_SPLIT_BWD = 'gather_fwd_split_bwd'
ALL2ALL_FWD_ALL2ALL_BWD = 'all2all_fwd_all2all_bwd'
SPLIT_FWD_GATHER_BWD = 'split_fwd_gather_bwd'
REDUCE_FWD_IDENTITY_BWD = 'all_reduce_fwd_identity_bwd'
ALLREDUCE_FWD_IDENTITY_BWD = 'all_reduce_fwd_identity_bwd'
IDENTITY_FWD_ALLREDUCE_BWD = 'identity_fwd_all_reduce_bwd'
@ -290,8 +290,8 @@ class CommSpec:
res_list.append(f"comm_pattern:SPLIT_FWD_GATHER_BWD, ")
res_list.append(f"shard_dim:{self.shard_dim}, ")
res_list.append(f"logical_process_axis:{self.logical_process_axis})")
elif self.comm_pattern == CollectiveCommPattern.REDUCE_FWD_IDENTITY_BWD:
res_list.append(f"comm_pattern:REDUCE_FWD_IDENTITY_BWD, ")
elif self.comm_pattern == CollectiveCommPattern.ALLREDUCE_FWD_IDENTITY_BWD:
res_list.append(f"comm_pattern:ALLREDUCE_FWD_IDENTITY_BWD, ")
res_list.append(f"logical_process_axis:{self.logical_process_axis})")
elif self.comm_pattern == CollectiveCommPattern.IDENTITY_FWD_ALLREDUCE_BWD:
res_list.append(f"comm_pattern:IDENTITY_FWD_ALLREDUCE_BWD, ")
@ -317,7 +317,7 @@ class CommSpec:
# all to all operation has same logical process axis as forward.
backward_communication_cost = self.device_mesh.all_to_all_cost(comm_size, self.logical_process_axis)
if self.comm_pattern == CollectiveCommPattern.REDUCE_FWD_IDENTITY_BWD:
if self.comm_pattern == CollectiveCommPattern.ALLREDUCE_FWD_IDENTITY_BWD:
forward_communication_cost = self.device_mesh.all_reduce_cost(comm_size, self.logical_process_axis)
backward_communication_cost = 0
@ -357,7 +357,7 @@ pattern_to_func_dict = {
CollectiveCommPattern.GATHER_FWD_SPLIT_BWD: gather_forward_split_backward,
CollectiveCommPattern.ALL2ALL_FWD_ALL2ALL_BWD: all_to_all,
CollectiveCommPattern.SPLIT_FWD_GATHER_BWD: split_forward_gather_backward,
CollectiveCommPattern.REDUCE_FWD_IDENTITY_BWD: reduce_input,
CollectiveCommPattern.ALLREDUCE_FWD_IDENTITY_BWD: reduce_input,
CollectiveCommPattern.IDENTITY_FWD_ALLREDUCE_BWD: reduce_grad,
}

4
tests/test_tensor/test_comm_spec_apply.py
View File

@ -136,7 +136,7 @@ def check_all_reduce_fwd(device_mesh, rank):
# device_mesh_shape: (2, 2)
sharding_spec = ShardingSpec(device_mesh, tensor_to_comm.shape, dim_partition_dict=dim_partition_dict)
comm_spec = CommSpec(CollectiveCommPattern.REDUCE_FWD_IDENTITY_BWD, sharding_spec, logical_process_axis=0)
comm_spec = CommSpec(CollectiveCommPattern.ALLREDUCE_FWD_IDENTITY_BWD, sharding_spec, logical_process_axis=0)
comm_spec.covert_spec_to_action(tensor_to_comm)
assert tensor_to_comm.equal(tensor_to_check)
@ -177,7 +177,7 @@ def check_all_reduce_in_flatten_device_mesh(device_mesh, rank):
sharding_spec = ShardingSpec(device_mesh, tensor_to_comm.shape, dim_partition_dict=dim_partition_dict)
# CommSpec:(comm_pattern:all_reduce, logical_process_axis:[0, 1])
comm_spec = CommSpec(CollectiveCommPattern.REDUCE_FWD_IDENTITY_BWD, sharding_spec, logical_process_axis=[0, 1])
comm_spec = CommSpec(CollectiveCommPattern.ALLREDUCE_FWD_IDENTITY_BWD, sharding_spec, logical_process_axis=[0, 1])
comm_spec.covert_spec_to_action(tensor_to_comm)
assert tensor_to_comm.equal(tensor_to_check)