Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Add multi-tensor hvd.grouped_allreduce API. (#2453)
Signed-off-by: Josh Romero <joshr@nvidia.com>
- Loading branch information
1 parent
c7848ca
commit 775959d
Showing
35 changed files
with
2,049 additions
and
274 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,82 @@ | ||
| // Copyright (c) 2020, NVIDIA CORPORATION. All rights reserved. | ||
| // | ||
| // Licensed under the Apache License, Version 2.0 (the "License"); | ||
| // you may not use this file except in compliance with the License. | ||
| // You may obtain a copy of the License at | ||
| // | ||
| // http://www.apache.org/licenses/LICENSE-2.0 | ||
| // | ||
| // Unless required by applicable law or agreed to in writing, software | ||
| // distributed under the License is distributed on an "AS IS" BASIS, | ||
| // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | ||
| // See the License for the specific language governing permissions and | ||
| // limitations under the License. | ||
| // ============================================================================= | ||
|
|
||
| #include "group_table.h" | ||
|
|
||
| #include <assert.h> | ||
|
|
||
| namespace horovod { | ||
| namespace common { | ||
|
|
||
| int32_t GroupTable::GetGroupIDFromTensorName(const std::string& tensor_name) const { | ||
| std::lock_guard<std::mutex> guard(mutex_); | ||
| auto it = tensor_name_to_id_.find(tensor_name); | ||
| if (it != tensor_name_to_id_.end()) | ||
| return it->second; | ||
| else { | ||
| return NULL_GROUP_ID; | ||
| } | ||
| } | ||
|
|
||
| const std::vector<std::string>& GroupTable::GetGroupTensorNames(int32_t group_id) const { | ||
| std::lock_guard<std::mutex> guard(mutex_); | ||
| return id_to_tensor_names_.at(group_id); | ||
| } | ||
|
|
||
| bool GroupTable::empty(void) const { | ||
| std::lock_guard<std::mutex> guard(mutex_); | ||
| return tensor_name_to_id_.empty(); | ||
| } | ||
|
|
||
| int32_t GroupTable::RegisterGroup(std::vector<std::string>&& tensor_names) { | ||
| std::lock_guard<std::mutex> guard(mutex_); | ||
|
|
||
| int32_t group_id; | ||
| if (!free_ids_.empty()) { | ||
| // Reuse old group_id | ||
| group_id = free_ids_.front(); | ||
| free_ids_.pop(); | ||
| } else { | ||
| // Create a new group_id | ||
| group_id = next_group_id_++; | ||
| } | ||
|
|
||
| for (auto& name : tensor_names) { | ||
| tensor_name_to_id_.emplace(name, group_id); | ||
| } | ||
| id_to_tensor_names_.emplace(group_id, std::move(tensor_names)); | ||
|
|
||
| return group_id; | ||
| } | ||
|
|
||
| void GroupTable::DeregisterGroups(const std::vector<std::string>& tensor_names) { | ||
| std::lock_guard<std::mutex> guard(mutex_); | ||
|
|
||
| for (auto& name : tensor_names) { | ||
| auto it = tensor_name_to_id_.find(name); | ||
| if (it != tensor_name_to_id_.end()) { | ||
| auto group_id = it->second; | ||
| for (auto& entry : id_to_tensor_names_[group_id]) { | ||
| tensor_name_to_id_.erase(entry); | ||
| } | ||
| id_to_tensor_names_.erase(group_id); | ||
|
|
||
| free_ids_.push(group_id); | ||
| } | ||
| } | ||
| } | ||
|
|
||
| } // namespace common | ||
| } // namespace horovod |
Oops, something went wrong.