[arm_ut_py] fix reduce_max compute error and add reduce_* ut_py

lite/backends/arm/math/reduce_max.h

@@ -83,7 +83,7 @@ inline void reduce_third_of_three(
     const T* src, T* dst, int first_in, int second_in, int third_in) {
   for (int i = 0; i < first_in; i++) {
     for (int j = 0; j < second_in; j++) {
-      dst[i * second_in + j] = src[i * second_in * third_in + j * second_in];
+      dst[i * second_in + j] = src[i * second_in * third_in + j * third_in];
       for (int k = 0; k < third_in; k++) {
         dst[i * second_in + j] =
             src[i * second_in * third_in + j * third_in + k] >

lite/tests/unittest_py/op/test_reduce_max_op.py

@@ -31,11 +31,16 @@
 class TestReduceMaxOp(AutoScanTest):
     def __init__(self, *args, **kwargs):
         AutoScanTest.__init__(self, *args, **kwargs)
+        self.enable_testing_on_place(
+            TargetType.ARM,
+            PrecisionType.FP32,
+            DataLayoutType.NCHW,
+            thread=[1, 4])
         self.enable_testing_on_place(
             TargetType.X86,
             PrecisionType.FP32,
             DataLayoutType.NCHW,
-            thread=[1, 2])
+            thread=[1, 4])
         opencl_places = [
             Place(TargetType.OpenCL, PrecisionType.FP16,
                   DataLayoutType.ImageDefault), Place(
@@ -69,13 +74,25 @@ def sample_program_configs(self, draw):
         in_shape = draw(
             st.lists(
                 st.integers(
-                    min_value=1, max_value=10), min_size=4, max_size=4))
+                    min_value=1, max_value=10), min_size=1, max_size=4))
         keep_dim = draw(st.booleans())
-        axis = draw(st.integers(min_value=-1, max_value=3))
+        axis_type = draw(st.sampled_from(["int", "list"]))
+        axis_int = draw(st.integers(min_value=-1, max_value=3))
+        axis_list = [0]
         assume(axis < len(in_shape))
+        if len(in_shape) == 2:
+            axis_list = draw(st.sampled_from([[0], [1]]))
+        elif len(in_shape) == 3:
+            axis_list = draw(st.sampled_from([[0], [1], [2]]))
+        elif len(in_shape) == 4:
+            axis_list = draw(
+                st.sampled_from([[0], [1], [2], [3], [0, 1], [1, 2], [2, 3]]))
+
+        if axis_type == "int":
+            axis = axis_int
+        else:
+            axis = axis_list
 
-        if isinstance(axis, int):
-            axis = [axis]
         reduce_all_data = True if axis == None or axis == [] else False
 
         def generate_input(*args, **kwargs):
@@ -113,18 +130,22 @@ def _teller2(program_config, predictor_config):
             in_shape = list(program_config.inputs["input_data"].shape)
             axis = program_config.ops[0].attrs["dim"]
             keep_dim = program_config.ops[0].attrs["keep_dim"]
+            if target_type == TargetType.OpenCL:
+                if len(in_shape) < 4:
+                    return True
             if target_type == TargetType.Metal:
-                if keep_dim == False or axis[0] != 1 or in_shape[0] != 1:
+                if keep_dim == False or axis[0] != 1 or in_shape[
+                        0] != 1 or len(in_shape) < 4:
                     return True
 
         self.add_ignore_check_case(
             _teller2, IgnoreReasons.PADDLELITE_NOT_SUPPORT,
-            "The op output has diff in a specific case on metal. We need to fix it as soon as possible."
+            "The op output has diff in a specific case on metal/opencl. We need to fix it as soon as possible."
         )
 
     def test(self, *args, **kwargs):
         target_str = self.get_target()
-        max_examples = 100
+        max_examples = 300
         if target_str == "Metal":
             # Make sure to generate enough valid cases for Metal
             max_examples = 3000

lite/tests/unittest_py/op/test_reduce_mean_op.py

@@ -64,14 +64,13 @@ def sample_program_configs(self, draw):
         axis_type = draw(st.sampled_from(["int", "list"]))
         axis_int = draw(st.integers(min_value=-1, max_value=3))
         axis_list = draw(
-            st.sampled_from([[2, 3], [1, 2], [0, 1], [1, 2, 3], [0, 1, 2]]))
+            st.sampled_from([[0], [1], [2], [3], [0, 1], [1, 2], [2, 3]]))
+        assume(axis < len(in_shape))
 
         if axis_type == "int":
             axis = axis_int
         else:
             axis = axis_list
-        if isinstance(axis, int):
-            axis = [axis]
         reduce_all_data = True if axis == None or axis == [] else False
 
         def generate_input(*args, **kwargs):

lite/tests/unittest_py/op/test_reduce_min_op.py

@@ -31,11 +31,16 @@
 class TestReduceMinOp(AutoScanTest):
     def __init__(self, *args, **kwargs):
         AutoScanTest.__init__(self, *args, **kwargs)
+        self.enable_testing_on_place(
+            TargetType.ARM,
+            PrecisionType.FP32,
+            DataLayoutType.NCHW,
+            thread=[1, 4])
         self.enable_testing_on_place(
             TargetType.X86,
             PrecisionType.FP32,
             DataLayoutType.NCHW,
-            thread=[1, 2])
+            thread=[1, 4])
 
     def is_program_valid(self,
                          program_config: ProgramConfig,
@@ -46,13 +51,24 @@ def sample_program_configs(self, draw):
         in_shape = draw(
             st.lists(
                 st.integers(
-                    min_value=1, max_value=10), min_size=4, max_size=4))
+                    min_value=1, max_value=10), min_size=1, max_size=4))
         keep_dim = draw(st.booleans())
-        axis = draw(st.integers(min_value=-1, max_value=3))
+        axis_type = draw(st.sampled_from(["int", "list"]))
+        axis_int = draw(st.integers(min_value=-1, max_value=3))
+        axis_list = [0]
         assume(axis < len(in_shape))
-
-        if isinstance(axis, int):
-            axis = [axis]
+        if len(in_shape) == 2:
+            axis_list = draw(st.sampled_from([[0], [1]]))
+        elif len(in_shape) == 3:
+            axis_list = draw(st.sampled_from([[0], [1], [2]]))
+        elif len(in_shape) == 4:
+            axis_list = draw(
+                st.sampled_from([[0], [1], [2], [3], [0, 1], [1, 2], [2, 3]]))
+
+        if axis_type == "int":
+            axis = axis_int
+        else:
+            axis = axis_list
         reduce_all_data = True if axis == None or axis == [] else False
 
         def generate_input(*args, **kwargs):

lite/tests/unittest_py/op/test_reduce_prod_op.py

@@ -0,0 +1,97 @@
+# Copyright (c) 2021 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 sys
+sys.path.append('../')
+
+from auto_scan_test import AutoScanTest, IgnoreReasons
+from program_config import TensorConfig, ProgramConfig, OpConfig, CxxConfig, TargetType, PrecisionType, DataLayoutType, Place
+import unittest
+
+import hypothesis
+from hypothesis import given, settings, seed, example, assume
+import hypothesis.strategies as st
+
+from functools import partial
+import numpy as np
+import argparse
+
+
+class TestReduceMeanOp(AutoScanTest):
+    def __init__(self, *args, **kwargs):
+        AutoScanTest.__init__(self, *args, **kwargs)
+        self.enable_testing_on_place(TargetType.X86, PrecisionType.FP32,
+                                     DataLayoutType.NCHW)
+        self.enable_testing_on_place(
+            TargetType.ARM,
+            PrecisionType.FP32,
+            DataLayoutType.NCHW,
+            thread=[1, 4])
+
+    def is_program_valid(self,
+                         program_config: ProgramConfig,
+                         predictor_config: CxxConfig) -> bool:
+        return True
+
+    def sample_program_configs(self, draw):
+        in_shape = draw(
+            st.lists(
+                st.integers(
+                    min_value=1, max_value=10), min_size=4, max_size=4))
+        keep_dim = draw(st.booleans())
+        axis_type = draw(st.sampled_from(["int", "list"]))
+        axis_int = draw(st.integers(min_value=-1, max_value=3))
+        axis_list = draw(
+            st.sampled_from([[0], [1], [2], [3], [0, 1], [1, 2], [2, 3]]))
+        assume(axis < len(in_shape))
+
+        if axis_type == "int":
+            axis = axis_int
+        else:
+            axis = axis_list
+        reduce_all_data = True if axis == None or axis == [] else False
+
+        def generate_input(*args, **kwargs):
+            return np.random.random(in_shape).astype(np.float32)
+
+        build_ops = OpConfig(
+            type="reduce_prod",
+            inputs={"X": ["input_data"], },
+            outputs={"Out": ["output_data"], },
+            attrs={
+                "dim": axis,
+                "keep_dim": keep_dim,
+                "reduce_all": reduce_all_data,
+            })
+        program_config = ProgramConfig(
+            ops=[build_ops],
+            weights={},
+            inputs={
+                "input_data": TensorConfig(data_gen=partial(generate_input)),
+            },
+            outputs=["output_data"])
+        return program_config
+
+    def sample_predictor_configs(self):
+        return self.get_predictor_configs(), ["reduce_prod"], (1e-2, 1e-2)
+
+    def add_ignore_pass_case(self):
+        pass
+
+    def test(self, *args, **kwargs):
+        self.run_and_statis(quant=False, max_examples=100)
+
+
+if __name__ == "__main__":
+    unittest.main(argv=[''])

lite/tests/unittest_py/op/test_reduce_sum_op.py

@@ -0,0 +1,97 @@
+# Copyright (c) 2021 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 sys
+sys.path.append('../')
+
+from auto_scan_test import AutoScanTest, IgnoreReasons
+from program_config import TensorConfig, ProgramConfig, OpConfig, CxxConfig, TargetType, PrecisionType, DataLayoutType, Place
+import unittest
+
+import hypothesis
+from hypothesis import given, settings, seed, example, assume
+import hypothesis.strategies as st
+
+from functools import partial
+import numpy as np
+import argparse
+
+
+class TestReduceMeanOp(AutoScanTest):
+    def __init__(self, *args, **kwargs):
+        AutoScanTest.__init__(self, *args, **kwargs)
+        self.enable_testing_on_place(TargetType.X86, PrecisionType.FP32,
+                                     DataLayoutType.NCHW)
+        self.enable_testing_on_place(
+            TargetType.ARM,
+            PrecisionType.FP32,
+            DataLayoutType.NCHW,
+            thread=[1, 4])
+
+    def is_program_valid(self,
+                         program_config: ProgramConfig,
+                         predictor_config: CxxConfig) -> bool:
+        return True
+
+    def sample_program_configs(self, draw):
+        in_shape = draw(
+            st.lists(
+                st.integers(
+                    min_value=1, max_value=10), min_size=4, max_size=4))
+        keep_dim = draw(st.booleans())
+        axis_type = draw(st.sampled_from(["int", "list"]))
+        axis_int = draw(st.integers(min_value=-1, max_value=3))
+        axis_list = draw(
+            st.sampled_from([[0], [1], [2], [3], [0, 1], [1, 2], [2, 3]]))
+        assume(axis < len(in_shape))
+
+        if axis_type == "int":
+            axis = axis_int
+        else:
+            axis = axis_list
+        reduce_all_data = True if axis == None or axis == [] else False
+
+        def generate_input(*args, **kwargs):
+            return np.random.random(in_shape).astype(np.float32)
+
+        build_ops = OpConfig(
+            type="reduce_sum",
+            inputs={"X": ["input_data"], },
+            outputs={"Out": ["output_data"], },
+            attrs={
+                "dim": axis,
+                "keep_dim": keep_dim,
+                "reduce_all": reduce_all_data,
+            })
+        program_config = ProgramConfig(
+            ops=[build_ops],
+            weights={},
+            inputs={
+                "input_data": TensorConfig(data_gen=partial(generate_input)),
+            },
+            outputs=["output_data"])
+        return program_config
+
+    def sample_predictor_configs(self):
+        return self.get_predictor_configs(), ["reduce_sum"], (1e-2, 1e-2)
+
+    def add_ignore_pass_case(self):
+        pass
+
+    def test(self, *args, **kwargs):
+        self.run_and_statis(quant=False, max_examples=100)
+
+
+if __name__ == "__main__":
+    unittest.main(argv=[''])