test_micro_batch_Interleaved

mindspore/ccsrc/frontend/parallel/graph_util/pipeline_split_utils.cc

@@ -545,7 +545,8 @@ AnfNodePtr GetPreNode(const AnfNodePtr &node) {
       continue;
     }
     (void)node_queue.erase(node_queue.begin());
-    if (!IsInEndNodeBlackList(cur_node) && cur_node->HasPrimalAttr(NEED_GRAD)) {
+    auto prim = GetCNodePrimitive(cur_node);
+    if (!IsInEndNodeBlackList(cur_node) && cur_node->HasPrimalAttr(NEED_GRAD) && !prim->HasAttr("realdiv_flag")) {
       MS_LOG(INFO) << "Pipeline End node: " << cur_node->DebugString();
       return cur_node;
     }

mindspore/nn/wrap/__init__.py

@@ -18,7 +18,7 @@ Wrap cells for networks.
 Use the Wrapper to combine the loss or build the training steps.
 """
 from .cell_wrapper import ForwardValueAndGrad, TrainOneStepCell, WithLossCell, WithGradCell, WithEvalCell, \
-     ParameterUpdate, GetNextSingleOp, VirtualDatasetCellTriple, PipelineCell
+     ParameterUpdate, GetNextSingleOp, VirtualDatasetCellTriple, MicroBatchInterleaved, PipelineCell
 from .loss_scale import TrainOneStepWithLossScaleCell, DynamicLossScaleUpdateCell, FixedLossScaleUpdateCell
 from .grad_reducer import DistributedGradReducer
 from ..layer.timedistributed import TimeDistributed
@@ -30,6 +30,7 @@ __all__ = [
     "TrainOneStepCell",
     "WithLossCell",
     "WithGradCell",
+    "MicroBatchInterleaved",
     "PipelineCell",
     "WithEvalCell",
     "GetNextSingleOp",

mindspore/nn/wrap/cell_wrapper.py

@@ -480,6 +480,7 @@ class MicroBatchInterleaved(Cell):
         self.network = network
         self.interleave_num = interleave_num
         self.interleave_inputs = nn.CellList()
+        self.realdiv = P.RealDiv().add_prim_attr("realdiv_flag", True)
         for _ in range(interleave_num):
             interleave_data = _MicroBatch(interleave_num)
             interleave_data.strided_slice.add_prim_attr("strided_slice_flag", True)
@@ -490,7 +491,7 @@ class MicroBatchInterleaved(Cell):
         for i in range(self.interleave_num):
             interleave_input = self.interleave_inputs[i](i, *inputs)
             output += self.network(*interleave_input)
-        return output / self.interleave_num
+        return self.realdiv(output, self.interleave_num)
 
 
 class PipelineCell(Cell):

tests/ut/python/parallel/test_micro_batch_interleaved.py

@@ -0,0 +1,71 @@
+# Copyright 2021 Huawei Technologies Co., Ltd
+#
+# 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.
+
+import numpy as np
+import mindspore as ms
+import mindspore.nn as nn
+from mindspore import Tensor, context
+from mindspore.ops import operations as P
+from mindspore.common.api import _cell_graph_executor
+from mindspore.nn.wrap.cell_wrapper import MicroBatchInterleaved
+
+
+class Net(nn.Cell):
+    def __init__(self, strategy1, strategy2):
+        super().__init__()
+        self.matmul1 = P.MatMul().shard(strategy1)
+        self.matmul2 = P.MatMul().shard(strategy2)
+
+    def construct(self, x, y, b):
+        out = self.matmul1(x, y)
+        out = self.matmul2(out, b)
+        return out
+
+
+class NetWithLoss(nn.Cell):
+    def __init__(self, network):
+        super(NetWithLoss, self).__init__()
+        self.loss = P.ReLU()
+        self.network = network
+
+    def construct(self, x, y, b):
+        predict = self.network(x, y, b)
+        return self.loss(predict)
+
+
+def compile_net(net, x, y, b):
+    net.set_auto_parallel()
+    net.set_train()
+    _cell_graph_executor.compile(net, x, y, b)
+
+
+def test_micro_batch_interleaved():
+    """
+    Feature: test MicroBatchInterleaved in auto parallel.
+    Description: net with MicroBatchInterleaved in semi auto parallel.
+    Expectation: compile done without error.
+    """
+    context.set_context(mode=context.GRAPH_MODE)
+    context.set_auto_parallel_context(parallel_mode="semi_auto_parallel")
+    context.set_auto_parallel_context(device_num=8, global_rank=0, gradients_mean=True)
+    strategy1 = ((4, 2), (2, 1))
+    strategy2 = ((2, 4), (4, 1))
+    micro_batch_interleaved = 2
+    net = MicroBatchInterleaved(NetWithLoss(Net(strategy1, strategy2)), micro_batch_interleaved)
+
+    x = Tensor(np.ones([128, 32]), dtype=ms.float32)
+    y = Tensor(np.ones([32 * micro_batch_interleaved, 64]), dtype=ms.float32)
+    b = Tensor(np.ones([64 * micro_batch_interleaved, 64]), dtype=ms.float32)
+
+    compile_net(net, x, y, b)

tests/ut/python/parallel/test_pipeline_split.py

@@ -21,7 +21,7 @@ from mindspore.ops import operations as P
 from mindspore.common.parameter import Parameter
 from mindspore.common.initializer import initializer
 from mindspore.train.model import Model
-from mindspore.nn.wrap.cell_wrapper import PipelineCell
+from mindspore.nn.wrap.cell_wrapper import PipelineCell, MicroBatchInterleaved
 
 
 class DatasetLenet():
@@ -263,3 +263,93 @@ def test_pipeline_split_shared_parameter_stage1_opt_shard():
     optimizer = nn.Lamb(params, learning_rate=0.01)
     model = Model(net, optimizer=optimizer)
     model.train(2, dataset, dataset_sink_mode=False)
+
+
+def test_pipeline_split_with_micro_batch_interleaved_stage0():
+    """
+    Feature: test PipelineSplit with MicroBatchInterleaved in auto parallel.
+    Description: net with MicroBatchInterleaved in semi auto parallel.
+    Expectation: success.
+    """
+    context.set_auto_parallel_context(device_num=8, global_rank=0, pipeline_stages=2)
+    context.set_auto_parallel_context(parallel_mode="semi_auto_parallel")
+    data = Tensor(np.ones([32, 64]), dtype=ms.float32)
+    label = Tensor(np.ones([64, 64]), dtype=ms.float32)
+    strategy1 = ((4, 1), (1, 1))
+    strategy2 = ((2, 1), (1, 1))
+    micro_batch_interleaved = 2
+    net = PipelineCell(MicroBatchInterleaved(PipelineSplit(strategy1, strategy2), micro_batch_interleaved), 4)
+    params = net.network.network.cell.block[0].trainable_params()
+    dataset = DatasetLenet(data, label, 3)
+    optimizer = nn.Lamb(params, learning_rate=0.01)
+    model = Model(net, optimizer=optimizer)
+    model.train(2, dataset, dataset_sink_mode=False)
+    for _, param in model._train_network.parameters_and_names():
+        assert param.name != "cell.block.1.param"
+        assert param.name != "cell.block.1.param1"
+
+
+def test_pipeline_split_with_micro_batch_interleaved_stage1():
+    """
+    Feature: test PipelineSplit with MicroBatchInterleaved in auto parallel.
+    Description: net with MicroBatchInterleaved in semi auto parallel.
+    Expectation: success.
+    """
+    context.set_auto_parallel_context(device_num=8, global_rank=4, pipeline_stages=2)
+    context.set_auto_parallel_context(parallel_mode="semi_auto_parallel")
+    data = Tensor(np.ones([32, 64]), dtype=ms.float32)
+    label = Tensor(np.ones([64, 64]), dtype=ms.float32)
+    strategy1 = ((4, 1), (1, 1))
+    strategy2 = ((2, 1), (1, 1))
+    micro_batch_interleaved = 2
+    net = PipelineCell(MicroBatchInterleaved(PipelineSplit(strategy1, strategy2), micro_batch_interleaved), 4)
+    params = net.network.network.cell.block[1].trainable_params()
+    dataset = DatasetLenet(data, label, 3)
+    optimizer = nn.Lamb(params, learning_rate=0.01)
+    model = Model(net, optimizer=optimizer)
+    model.train(2, dataset, dataset_sink_mode=False)
+    for _, param in model._train_network.parameters_and_names():
+        assert param.name != "cell.block.0.param"
+        assert param.name != "cell.block.0.param1"
+
+
+def test_pipeline_split_shared_parameter_with_micro_batch_interleaved_stage0_opt_shard():
+    """
+    Feature: test PipelineSplitSharedParameter with MicroBatchInterleaved in auto parallel.
+    Description: net with MicroBatchInterleaved in semi auto parallel.
+    Expectation: success.
+    """
+    context.set_auto_parallel_context(device_num=8, global_rank=0, pipeline_stages=2, enable_parallel_optimizer=True)
+    context.set_auto_parallel_context(parallel_mode="semi_auto_parallel")
+    data = Tensor(np.ones([32, 64]), dtype=ms.float32)
+    label = Tensor(np.ones([64, 64]), dtype=ms.float32)
+    strategy1 = ((4, 1), (1, 1))
+    strategy2 = ((2, 1), (1, 1))
+    micro_batch_interleaved = 2
+    net = PipelineCell(MicroBatchInterleaved(PipelineSplit2(strategy1, strategy2), micro_batch_interleaved), 4)
+    params = net.network.network.cell.block[0].trainable_params()
+    dataset = DatasetLenet(data, label, 3)
+    optimizer = nn.Lamb(params, learning_rate=0.01)
+    model = Model(net, optimizer=optimizer)
+    model.train(2, dataset, dataset_sink_mode=False)
+
+
+def test_pipeline_split_shared_parameter_with_micro_batch_interleaved_stage1_opt_shard():
+    """
+    Feature: test PipelineSplitSharedParameter with MicroBatchInterleaved in auto parallel.
+    Description: net with MicroBatchInterleaved in semi auto parallel.
+    Expectation: success.
+    """
+    context.set_auto_parallel_context(device_num=8, global_rank=4, pipeline_stages=2, enable_parallel_optimizer=True)
+    context.set_auto_parallel_context(parallel_mode="semi_auto_parallel")
+    data = Tensor(np.ones([32, 64]), dtype=ms.float32)
+    label = Tensor(np.ones([64, 64]), dtype=ms.float32)
+    strategy1 = ((4, 1), (1, 1))
+    strategy2 = ((2, 1), (1, 1))
+    micro_batch_interleaved = 2
+    net = PipelineCell(MicroBatchInterleaved(PipelineSplit2(strategy1, strategy2), micro_batch_interleaved), 4)
+    params = net.network.network.cell.block[1].trainable_params()
+    dataset = DatasetLenet(data, label, 3)
+    optimizer = nn.Lamb(params, learning_rate=0.01)
+    model = Model(net, optimizer=optimizer)
+    model.train(2, dataset, dataset_sink_mode=False)