caoying03
8 years ago
commit
25083de998
@ -0,0 +1,15 @@
|
|||||||
|
#!/bin/bash
|
||||||
|
set -e
|
||||||
|
|
||||||
|
readonly VERSION="3.8"
|
||||||
|
|
||||||
|
version=$(clang-format -version)
|
||||||
|
|
||||||
|
if ! [[ $version == *"$VERSION"* ]]; then
|
||||||
|
echo "clang-format version check failed."
|
||||||
|
echo "a version contains '$VERSION' is needed, but get '$version'"
|
||||||
|
echo "you can install the right version, and make an soft-link to '\$PATH' env"
|
||||||
|
exit -1
|
||||||
|
fi
|
||||||
|
|
||||||
|
clang-format $@
|
||||||
@ -0,0 +1,101 @@
|
|||||||
|
# Alalysis of large model distributed training in Paddle
|
||||||
|
|
||||||
|
***NOTE: This is only some note for how we implemeted this scheme in V1, not a new design.***
|
||||||
|
|
||||||
|
## What is it
|
||||||
|
|
||||||
|
We often encounter cases that the embedding layer parameters(sparse) are so large that we can not store it in the trainer's memory when training. So we need to put them to several servers, and fetch them row by row instead of fetch all of the parameters.
|
||||||
|
|
||||||
|
## How to use
|
||||||
|
|
||||||
|
Specify command-line argument like `--loadsave_parameters_in_pserver=true --ports_num_for_sparse=1 --use_old_updater=1` when starting the paddle trainer. And also add something like `--ports_num_for_sparse=1 --pserver_num_threads=5` when starting pserver processes.
|
||||||
|
|
||||||
|
Accrodingly, configure your embedding layers like:
|
||||||
|
|
||||||
|
```python
|
||||||
|
SPARSE_REMOTE=True
|
||||||
|
|
||||||
|
w1 = data_layer(name="w1", size=dict_size)
|
||||||
|
emb1 = embedding_layer(input=w1, size=32, param_attr=ParameterAttribute(sparse_update=SPARSE_REMOTE))
|
||||||
|
w2 = data_layer(name="w2", size=dict_size)
|
||||||
|
emb2 = embedding_layer(input=w2, size=32, param_attr=ParameterAttribute(sparse_update=SPARSE_REMOTE))
|
||||||
|
...
|
||||||
|
```
|
||||||
|
|
||||||
|
## Implementation details
|
||||||
|
|
||||||
|
```c++
|
||||||
|
enum MatType {
|
||||||
|
MAT_NORMAL,
|
||||||
|
MAT_NORMAL_SHARED,
|
||||||
|
MAT_VALUE_SHARED,
|
||||||
|
MAT_SPARSE_ROW_IDS,
|
||||||
|
MAT_SPARSE_ROW_AUTO_GROW,
|
||||||
|
MAT_CACHE_ROW,
|
||||||
|
MAT_SPARSE_ROW,
|
||||||
|
MAT_SPARSE_ROW_PREFETCH,
|
||||||
|
MAT_SPARSE_ROW_PREFETCH_FULL_SIZE,
|
||||||
|
};
|
||||||
|
```
|
||||||
|
|
||||||
|
`MAT_SPARSE_ROW_PREFETCH` is what we use when configured to fetch only row of matrix when training.
|
||||||
|
|
||||||
|
In `trainer_internal.cpp:L93 trainOneBatch`:
|
||||||
|
|
||||||
|
```c++
|
||||||
|
if (config_->getOptConfig().use_sparse_remote_updater()) {
|
||||||
|
REGISTER_TIMER("prefetch");
|
||||||
|
gradientMachine_->prefetch(inArgs);
|
||||||
|
parameterUpdater_->getParametersRemote();
|
||||||
|
}
|
||||||
|
```
|
||||||
|
|
||||||
|
When doing actual network forward and backward, at the beginning of each batch, the trainer will try to download one row of data from pserver.
|
||||||
|
|
||||||
|
In `trainer/RemoteParameterUpdater.cpp`: `parameterUpdater_->getParametersRemote();`:
|
||||||
|
|
||||||
|
```c++
|
||||||
|
if (fullSize) {
|
||||||
|
...
|
||||||
|
} else {
|
||||||
|
getParams = [&] {
|
||||||
|
parameterClient_->getParameterSparse(
|
||||||
|
/* recvParameterType= */ PARAMETER_VALUE, sendBackParameterType);
|
||||||
|
};
|
||||||
|
applyL1 = [](Parameter& para, real decayRate) {
|
||||||
|
para.getMat(PARAMETER_VALUE)->applyL1(/*lr=*/1.0f, decayRate);
|
||||||
|
};
|
||||||
|
}
|
||||||
|
```
|
||||||
|
|
||||||
|
Calling `parameterClient_->getParameterSparse` will do remote call to pserver's `getParameterSparse`:
|
||||||
|
|
||||||
|
```c++
|
||||||
|
void ParameterServer2::getParameterSparse(const SendParameterRequest& request,
|
||||||
|
std::vector<Buffer>& inputBuffers,
|
||||||
|
SendParameterResponse* response,
|
||||||
|
std::vector<Buffer>* outputBuffers) {
|
||||||
|
(void)inputBuffers;
|
||||||
|
auto& buffer = *readWriteBuffer_;
|
||||||
|
size_t numReals = 0;
|
||||||
|
for (const auto& block : request.blocks()) {
|
||||||
|
numReals += getParameterConfig(block).dims(1);
|
||||||
|
}
|
||||||
|
buffer.resize(numReals);
|
||||||
|
|
||||||
|
VLOG(3) << "pserver: getParameterSparse, numReals=" << numReals;
|
||||||
|
|
||||||
|
ReadLockGuard guard(parameterMutex_);
|
||||||
|
size_t offset = 0;
|
||||||
|
for (const auto& block : request.blocks()) {
|
||||||
|
size_t width = getParameterConfig(block).dims(1);
|
||||||
|
Buffer buf = {buffer.data() + offset, width};
|
||||||
|
int type = request.send_back_parameter_type();
|
||||||
|
sendBackParameterSparse(block, type, response, &buf, width, outputBuffers);
|
||||||
|
offset += width;
|
||||||
|
}
|
||||||
|
}
|
||||||
|
```
|
||||||
|
|
||||||
|
`getParameterConfig(block).dims(1)` returns the width of the current "parameter block"(a shard of parameter object),
|
||||||
|
then `getParameterSparse` remote call returns only one row of data to the client.
|
||||||
@ -1,28 +0,0 @@
|
|||||||
/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserve.
|
|
||||||
|
|
||||||
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. */
|
|
||||||
|
|
||||||
syntax = "proto2";
|
|
||||||
package paddle.framework;
|
|
||||||
|
|
||||||
// Attribute Type for paddle's Op.
|
|
||||||
// Op contains many attributes. Each type of attributes could be different.
|
|
||||||
// The AttrType will be shared between AttrDesc and AttrProto.
|
|
||||||
enum AttrType {
|
|
||||||
INT = 0;
|
|
||||||
FLOAT = 1;
|
|
||||||
STRING = 2;
|
|
||||||
INTS = 3;
|
|
||||||
FLOATS = 4;
|
|
||||||
STRINGS = 5;
|
|
||||||
}
|
|
||||||
File diff suppressed because it is too large
Load Diff
@ -0,0 +1,82 @@
|
|||||||
|
/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserve.
|
||||||
|
|
||||||
|
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. */
|
||||||
|
|
||||||
|
syntax = "proto2";
|
||||||
|
package paddle.framework;
|
||||||
|
|
||||||
|
enum AttrType {
|
||||||
|
INT = 0;
|
||||||
|
FLOAT = 1;
|
||||||
|
STRING = 2;
|
||||||
|
INTS = 3;
|
||||||
|
FLOATS = 4;
|
||||||
|
STRINGS = 5;
|
||||||
|
}
|
||||||
|
|
||||||
|
// OpDesc describes an instance of a C++ framework::OperatorBase
|
||||||
|
// derived class type.
|
||||||
|
message OpDesc {
|
||||||
|
|
||||||
|
message Attr {
|
||||||
|
required string name = 1;
|
||||||
|
required AttrType type = 2;
|
||||||
|
optional int32 i = 3;
|
||||||
|
optional float f = 4;
|
||||||
|
optional string s = 5;
|
||||||
|
repeated int32 ints = 6;
|
||||||
|
repeated float floats = 7;
|
||||||
|
repeated string strings = 8;
|
||||||
|
};
|
||||||
|
|
||||||
|
message Var {
|
||||||
|
required string parameter = 1;
|
||||||
|
repeated string arguments = 2;
|
||||||
|
};
|
||||||
|
|
||||||
|
required string type = 3;
|
||||||
|
repeated Var inputs = 1;
|
||||||
|
repeated Var outputs = 2;
|
||||||
|
repeated Attr attrs = 4;
|
||||||
|
};
|
||||||
|
|
||||||
|
// OpProto describes a C++ framework::OperatorBase derived class.
|
||||||
|
message OpProto {
|
||||||
|
|
||||||
|
// VarProto describes the C++ type framework::Variable.
|
||||||
|
message Var {
|
||||||
|
required string name = 1;
|
||||||
|
required string comment = 2;
|
||||||
|
|
||||||
|
optional bool duplicable = 3 [ default = false ];
|
||||||
|
optional bool intermediate = 4 [ default = false ];
|
||||||
|
optional bool not_in_gradient = 5 [ default = false ];
|
||||||
|
}
|
||||||
|
|
||||||
|
// AttrProto describes the C++ type Attribute.
|
||||||
|
message Attr {
|
||||||
|
required string name = 1;
|
||||||
|
required AttrType type = 2;
|
||||||
|
required string comment = 3;
|
||||||
|
// If that attribute is generated, it means the Paddle third
|
||||||
|
// language binding has responsibility to fill that
|
||||||
|
// attribute. End-User should not set that attribute.
|
||||||
|
optional bool generated = 4 [ default = false ];
|
||||||
|
}
|
||||||
|
|
||||||
|
required string type = 1;
|
||||||
|
repeated Var inputs = 2;
|
||||||
|
repeated Var outputs = 3;
|
||||||
|
repeated Attr attrs = 4;
|
||||||
|
required string comment = 5;
|
||||||
|
}
|
||||||
Some files were not shown because too many files have changed in this diff Show More
Loading…
Reference in new issue