[AutoParallel] fix input order (#49329)

* fix input order

* add unittest

* update cmakelist

python/paddle/distributed/auto_parallel/dist_saver.py

@@ -202,7 +202,9 @@ class DistributedSaver:
             if var_name not in used_outputs:
                 fetch_vars_names.pop(idx)
 
-        dist_feed_vars = [global_block.vars[name] for name in feed_vars_names]
+        dist_feed_vars = list(
+            reversed([global_block.vars[name] for name in feed_vars_names])
+        )
         dist_fetch_vars = [global_block.vars[name] for name in fetch_vars_names]
 
         dist_filename = filename + "_dist" + str(rank_id)

python/paddle/fluid/tests/unittests/auto_parallel/CMakeLists.txt

@@ -128,4 +128,5 @@ if(WITH_DISTRIBUTE AND WITH_GPU)
   py_test_modules(test_cluster_partition MODULES test_cluster_partition)
   py_test_modules(test_convert_to_process_meshes MODULES
                   test_convert_to_process_meshes)
+  py_test_modules(test_dist_saver MODULES test_dist_saver)
 endif()

python/paddle/fluid/tests/unittests/auto_parallel/test_dist_saver.py

+# Copyright (c) 2022 PaddlePaddle Authors. 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.
+
+import os
+import tempfile
+import unittest
+
+import paddle
+import paddle.nn as nn
+import paddle.nn.functional as F
+from paddle.distributed.fleet import auto
+
+paddle.enable_static()
+
+
+class MLPLayer(nn.Layer):
+    def __init__(
+        self,
+        hidden_size=1024,
+        intermediate_size=4 * 1024,
+        dropout_ratio=0.1,
+        initializer_range=0.02,
+    ):
+        super().__init__()
+        d_model = hidden_size
+        dim_feedforward = intermediate_size
+        weight_attr = paddle.ParamAttr(
+            initializer=nn.initializer.Normal(mean=0.0, std=initializer_range)
+        )
+        bias_attr = None
+
+        self.linear0 = nn.Linear(
+            d_model, dim_feedforward, weight_attr, bias_attr=bias_attr
+        )
+        self.linear1 = nn.Linear(
+            dim_feedforward, d_model, weight_attr, bias_attr=bias_attr
+        )
+        self.linear2 = nn.Linear(d_model, 1, weight_attr, bias_attr=bias_attr)
+        self.norm = nn.LayerNorm(d_model, epsilon=1e-5)
+        self.dropout = nn.Dropout(dropout_ratio, mode="upscale_in_train")
+
+    def forward(self, input0, input1):
+        out = self.norm(input0)
+        out = self.linear0(out)
+        out = out + input1
+        out = F.gelu(out, approximate=True)
+        out = self.linear1(out)
+        out = self.dropout(out)
+        out = self.linear2(out)
+        return out
+
+
+class TestDistSaver(unittest.TestCase):
+    def test_dist_saver(self):
+
+        mlp = MLPLayer()
+        loss = paddle.nn.CrossEntropyLoss()
+        optimizer = paddle.optimizer.Adam(
+            learning_rate=0.00001,
+            beta1=0.9,
+            beta2=0.999,
+            epsilon=1e-08,
+            grad_clip=None,
+        )
+        metric = paddle.metric.Accuracy()
+
+        strategy = auto.Strategy()
+        strategy.auto_mode = "semi"
+
+        engine = auto.Engine(mlp, loss, optimizer, metric, strategy=strategy)
+
+        inputs_spec = [
+            paddle.static.InputSpec(
+                shape=[2, 1024], dtype="float32", name="input0"
+            ),
+            paddle.static.InputSpec(
+                shape=[2, 4096], dtype="float32", name="input1"
+            ),
+        ]
+
+        engine.prepare(inputs_spec, mode="predict")
+        temp_dir = tempfile.TemporaryDirectory()
+        model_filename = os.path.join(temp_dir.name, 'mlp')
+        engine.save(model_filename, training=False)
+
+        with open(model_filename + "_dist0.pdmodel", 'rb') as f:
+            data = f.read()
+
+        program = paddle.static.io.deserialize_program(data)
+
+        input_vars = []
+        for op in program.global_block().ops:
+            if op.type == "feed":
+                input_vars.append(op.output_arg_names[0])
+            else:
+                break
+
+        assert input_vars == ["input0", "input1"]
+
+
+if __name__ == "__main__":
+    unittest.main()