convert SRGAN model test

Author: hanjr92

tests/export_application/test_export_srgan.py

@@ -23,7 +23,10 @@ def convert_padding(padding, input_shape, output_shape, kernel_shape, strides, d
                 output_shape = make_shape_channels_first(output_shape)
 
             if any(input_shape[i + 2] == -1 or output_shape[i + 2] == -1 for i in range(spatial)):
-                return  "SAME_UPPER"
+
+                auto_pad = "SAME_UPPER"
+
+                return  auto_pad
 
             for i in range(spatial):
                 pad = (
@@ -38,14 +41,14 @@ def convert_padding(padding, input_shape, output_shape, kernel_shape, strides, d
             return pads
 
         elif padding == "VALID":
-            return "VALID"
+            auto_pad = "VALID"
+            return auto_pad
     elif isinstance(padding, int):
         pads = [padding] * spatial * 2
         return pads
     elif isinstance(padding, tuple):
         return list(padding) * 2
 
-
 def convert_w(w, data_format, spatial, w_name):
     w = tlx.convert_to_numpy(w)
     if tlx.BACKEND == 'tensorflow':
@@ -62,6 +65,9 @@ def convert_w(w, data_format, spatial, w_name):
             return numpy_helper.from_array(w, name=w_name)
     return numpy_helper.from_array(w, name=w_name)
 
+def convert_b(b, b_name):
+    b = tlx.convert_to_numpy(b)
+    return numpy_helper.from_array(b, name=b_name)
 
 def convert_tlx_relu(inputs, outputs, act = None):
     opsets = OpMapper.OPSETS['ReLU']

tlx2onnx/common/preprocessing.py

@@ -1,4 +1,5 @@
 #! /usr/bin/python
 # -*- coding: utf-8 -*-
 
-from . import matmul
+from . import matmul
+from . import add

tlx2onnx/op_mapper/math/__init__.py

@@ -0,0 +1,30 @@
+#! /usr/bin/python
+# -*- coding: utf-8 -*-
+
+from onnx import helper, TensorProto
+from ..op_mapper import OpMapper
+from ...common import make_node, transpose_shape
+from tlx2onnx.op_mapper.datatype_mapping import NP_TYPE_TO_TENSOR_TYPE
+
+@OpMapper('Add')
+class Add():
+    # supports v7-v12
+
+    @classmethod
+    def version_7(cls, node, **kwargs):
+        onnx_node = []
+        onnx_value = []
+        onnx_init = []
+
+        x_name = node['in_nodes_name'][0]
+        y_name = node['in_nodes_name'][1]
+        out_name = node['out_nodes_name'][0]
+        # x_shape = node['in_tensors'][0]
+        # y_shape = node['in_tensors'][1]
+        # out_shape = node['out_tensors'][0]
+
+        op_type = 'Add'
+        add_node, _ = make_node(op_type, inputs=[x_name, y_name], outputs=[out_name])
+        onnx_node.append(add_node)
+        return onnx_node, onnx_value, onnx_init
+

tlx2onnx/op_mapper/math/add.py

@@ -23,5 +23,5 @@
 from .stack import *
 from .subpixelconv import *
 from .mask_conv import *
-from .group_conv import *
+from .groupconv import *
 

tlx2onnx/op_mapper/nn/__init__.py

@@ -7,64 +7,8 @@
 from tlx2onnx.op_mapper.datatype_mapping import NP_TYPE_TO_TENSOR_TYPE
 from tlx2onnx.op_mapper.op_mapper import OpMapper
 from tlx2onnx.common import make_node
-from tlx2onnx.common import make_shape_channels_first, get_channels_first_permutation,tlx_act_2_onnx,get_channels_last_permutation
-
-
-def convert_padding(padding, input_shape, output_shape, kernel_shape, strides, dilations, spatial, data_format):
-    if isinstance(padding, str):
-        if padding == "SAME":
-            pads = [0] * (spatial * 2)
-            if data_format == "channels_last":
-                input_shape = make_shape_channels_first(input_shape)
-                output_shape = make_shape_channels_first(output_shape)
-
-            if any(input_shape[i + 2] == -1 or output_shape[i + 2] == -1 for i in range(spatial)):
-
-                auto_pad = "SAME_UPPER"
-
-                return  auto_pad
-
-            for i in range(spatial):
-                pad = (
-                    (output_shape[i + 2] - 1) * strides[i]
-                    + dilations[i] * (kernel_shape[i] - 1) + 1
-                    - input_shape[i + 2]
-                )
-                pad = max(pad, 0)
-                pads[i] = pad // 2
-                pads[i + spatial] = pad - pad // 2
-
-            return pads
-
-        elif padding == "VALID":
-            auto_pad = "VALID"
-            return auto_pad
-    elif isinstance(padding, int):
-        pads = [padding] * spatial * 2
-        return pads
-    elif isinstance(padding, tuple):
-        return list(padding) * 2
-
-def convert_w(w, data_format, spatial, w_name):
-    w = tlx.convert_to_numpy(w)
-    if tlx.BACKEND == 'tensorflow':
-        if spatial == 2:
-            w = np.transpose(w, axes=[3, 2, 0, 1])
-        elif spatial == 1:
-            w = np.transpose(w, axes=[2, 1, 0])
-        elif spatial == 3:
-            w = np.transpose(w, axes=[4, 3, 0, 1, 2])
-        return numpy_helper.from_array(w, name=w_name)
-    elif tlx.BACKEND == 'mindspore':
-        if spatial == 2 and data_format == 'channels_last':
-            w = np.transpose(w, axes=[3, 0, 1, 2])
-            return numpy_helper.from_array(w, name=w_name)
-    return numpy_helper.from_array(w, name=w_name)
-
-def convert_b(b, b_name):
-    b = tlx.convert_to_numpy(b)
-    return numpy_helper.from_array(b, name=b_name)
-
+from tlx2onnx.common import make_shape_channels_first, get_channels_first_permutation,get_channels_last_permutation
+from tlx2onnx.common import convert_padding, convert_w, tlx_act_2_onnx, convert_b
 
 @OpMapper(["Conv1d", "Conv2d", "Conv3d"])
 class Conv():

tlx2onnx/op_mapper/nn/conv.py

@@ -7,64 +7,8 @@
 from tlx2onnx.op_mapper.datatype_mapping import NP_TYPE_TO_TENSOR_TYPE
 from tlx2onnx.op_mapper.op_mapper import OpMapper
 from tlx2onnx.common import make_node
-from tlx2onnx.common import make_shape_channels_first, get_channels_first_permutation,tlx_act_2_onnx,get_channels_last_permutation
-
-
-def convert_padding(padding, input_shape, output_shape, kernel_shape, strides, dilations, spatial, data_format):
-    if isinstance(padding, str):
-        if padding == "SAME":
-            pads = [0] * (spatial * 2)
-            if data_format == "channels_last":
-                input_shape = make_shape_channels_first(input_shape)
-                output_shape = make_shape_channels_first(output_shape)
-
-            if any(input_shape[i + 2] == -1 or output_shape[i + 2] == -1 for i in range(spatial)):
-
-                auto_pad = "SAME_UPPER"
-
-                return  auto_pad
-
-            for i in range(spatial):
-                pad = (
-                    (output_shape[i + 2] - 1) * strides[i]
-                    + dilations[i] * (kernel_shape[i] - 1) + 1
-                    - input_shape[i + 2]
-                )
-                pad = max(pad, 0)
-                pads[i] = pad // 2
-                pads[i + spatial] = pad - pad // 2
-
-            return pads
-
-        elif padding == "VALID":
-            auto_pad = "VALID"
-            return auto_pad
-    elif isinstance(padding, int):
-        pads = [padding] * spatial * 2
-        return pads
-    elif isinstance(padding, tuple):
-        return list(padding) * 2
-
-def convert_w(w, data_format, spatial, w_name):
-    w = tlx.convert_to_numpy(w)
-    if tlx.BACKEND == 'tensorflow':
-        if spatial == 2:
-            w = np.transpose(w, axes=[3, 2, 0, 1])
-        elif spatial == 1:
-            w = np.transpose(w, axes=[2, 1, 0])
-        elif spatial == 3:
-            w = np.transpose(w, axes=[4, 3, 0, 1, 2])
-        return numpy_helper.from_array(w, name=w_name)
-    elif tlx.BACKEND == 'mindspore':
-        if spatial == 2 and data_format == 'channels_last':
-            w = np.transpose(w, axes=[3, 0, 1, 2])
-            return numpy_helper.from_array(w, name=w_name)
-    return numpy_helper.from_array(w, name=w_name)
-
-def convert_b(b, b_name):
-    b = tlx.convert_to_numpy(b)
-    return numpy_helper.from_array(b, name=b_name)
-
+from tlx2onnx.common import make_shape_channels_first, get_channels_first_permutation,get_channels_last_permutation
+from tlx2onnx.common import convert_padding, convert_w, tlx_act_2_onnx, convert_b
 
 @OpMapper(["GroupConv2d"])
 class Conv():