mindspore 2.2.0__cp38-cp38-manylinux1_x86_64.whl → 2.2.11__cp38-cp38-manylinux1_x86_64.whl

mindspore/lib/plugin/ascend/custom_aicpu_ops/op_impl/cpu/config/cust_aicpu_kernel.json

@@ -3292,6 +3292,41 @@
             "type":"DT_DOUBLE,DT_FLOAT,DT_FLOAT16"
         }
     },
+    "CustLinearSumAssignment":{
+        "input0":{
+            "name":"cost_matrix",
+            "type":"DT_FLOAT16,DT_FLOAT,DT_DOUBLE,DT_INT16,DT_INT32,DT_INT64,DT_INT8,DT_UINT16,DT_UINT32,DT_UINT64,DT_UINT8,DT_BOOL"
+        },
+        "input1":{
+            "name":"dimension_limit",
+            "type":"DT_INT64"
+        },
+        "input2":{
+            "name":"maximize",
+            "type":"DT_BOOL"
+        },
+        "opInfo":{
+            "computeCost":"100",
+            "engine":"DNN_VM_AICPU",
+            "flagAsync":"False",
+            "flagPartial":"False",
+            "formatAgnostic":"False",
+            "functionName":"RunCpuKernel",
+            "kernelSo":"libcust_cpu_kernels.so",
+            "opKernelLib":"CUSTAICPUKernel",
+            "opsFlag":"OPS_FLAG_OPEN",
+            "subTypeOfInferShape":"1",
+            "userDefined":"True"
+        },
+        "output0":{
+            "name":"row_ind",
+            "type":"DT_INT64"
+        },
+        "output1":{
+            "name":"col_ind",
+            "type":"DT_INT64"
+        }
+    },
     "CustListDiff":{
         "input0":{
             "name":"x",
@@ -3888,42 +3923,6 @@
             "type":"DT_FLOAT,DT_FLOAT16,DT_DOUBLE"
         }
     },
-    "CustMultinomial":{
-        "input0":{
-            "name":"input",
-            "type":"DT_DOUBLE,DT_FLOAT,DT_FLOAT16"
-        },
-        "input1":{
-            "name":"num_sample",
-            "type":"DT_INT32"
-        },
-        "input2":{
-            "name":"count",
-            "type":"DT_UINT64"
-        },
-        "input3":{
-            "name":"state",
-            "type":"DT_UINT64"
-        },
-        "opInfo":{
-            "computeCost":"100",
-            "engine":"DNN_VM_AICPU",
-            "flagAsync":"False",
-            "flagPartial":"False",
-            "formatAgnostic":"False",
-            "functionName":"RunCpuKernel",
-            "kernelSo":"libcust_cpu_kernels.so",
-            "opKernelLib":"CUSTAICPUKernel",
-            "opsFlag":"OPS_FLAG_CLOSE",
-            "subTypeOfInferShape":"1",
-            "userDefined":"True",
-            "workspaceSize":"1024"
-        },
-        "output0":{
-            "name":"output",
-            "type":"DT_INT32,DT_INT64"
-        }
-    },
     "CustMvlgamma":{
         "input0":{
             "name":"x",
@@ -4599,7 +4598,7 @@
     "CustResizeBicubic":{
         "input0":{
             "name":"images",
-            "type":"DT_INT8,DT_UINT8,DT_INT16,DT_UINT16,DT_INT32,DT_INT64,DT_FLOAT16,DT_FLOAT,DT_DOUBLE"
+            "type":"DT_FLOAT16,DT_FLOAT,DT_DOUBLE"
         },
         "input1":{
             "name":"size",
@@ -4621,13 +4620,13 @@
         },
         "output0":{
             "name":"y",
-            "type":"DT_FLOAT"
+            "type":"DT_FLOAT16,DT_FLOAT,DT_DOUBLE"
         }
     },
     "CustResizeBicubicGrad":{
         "input0":{
             "name":"grads",
-            "type":"DT_FLOAT"
+            "type":"DT_FLOAT,DT_DOUBLE"
         },
         "input1":{
             "name":"original_image",
@@ -4671,7 +4670,7 @@
             "kernelSo":"libcust_cpu_kernels.so",
             "opKernelLib":"CUSTAICPUKernel",
             "opsFlag":"OPS_FLAG_CLOSE",
-            "subTypeOfInferShape":"4",
+            "subTypeOfInferShape":"2",
             "userDefined":"True",
             "workspaceSize":"1024"
         },
@@ -6640,6 +6639,41 @@
             "type":"DT_DOUBLE,DT_FLOAT,DT_FLOAT16"
         }
     },
+    "LinearSumAssignment":{
+        "input0":{
+            "name":"cost_matrix",
+            "type":"DT_FLOAT16,DT_FLOAT,DT_DOUBLE,DT_INT16,DT_INT32,DT_INT64,DT_INT8,DT_UINT16,DT_UINT32,DT_UINT64,DT_UINT8,DT_BOOL"
+        },
+        "input1":{
+            "name":"dimension_limit",
+            "type":"DT_INT64"
+        },
+        "input2":{
+            "name":"maximize",
+            "type":"DT_BOOL"
+        },
+        "opInfo":{
+            "computeCost":"100",
+            "engine":"DNN_VM_AICPU",
+            "flagAsync":"False",
+            "flagPartial":"False",
+            "formatAgnostic":"False",
+            "functionName":"RunCpuKernel",
+            "kernelSo":"libcust_cpu_kernels.so",
+            "opKernelLib":"CUSTAICPUKernel",
+            "opsFlag":"OPS_FLAG_OPEN",
+            "subTypeOfInferShape":"1",
+            "userDefined":"True"
+        },
+        "output0":{
+            "name":"row_ind",
+            "type":"DT_INT64"
+        },
+        "output1":{
+            "name":"col_ind",
+            "type":"DT_INT64"
+        }
+    },
     "ListDiff":{
         "input0":{
             "name":"x",
@@ -7236,42 +7270,6 @@
             "type":"DT_FLOAT,DT_FLOAT16,DT_DOUBLE"
         }
     },
-    "Multinomial":{
-        "input0":{
-            "name":"input",
-            "type":"DT_DOUBLE,DT_FLOAT,DT_FLOAT16"
-        },
-        "input1":{
-            "name":"num_sample",
-            "type":"DT_INT32"
-        },
-        "input2":{
-            "name":"count",
-            "type":"DT_UINT64"
-        },
-        "input3":{
-            "name":"state",
-            "type":"DT_UINT64"
-        },
-        "opInfo":{
-            "computeCost":"100",
-            "engine":"DNN_VM_AICPU",
-            "flagAsync":"False",
-            "flagPartial":"False",
-            "formatAgnostic":"False",
-            "functionName":"RunCpuKernel",
-            "kernelSo":"libcust_cpu_kernels.so",
-            "opKernelLib":"CUSTAICPUKernel",
-            "opsFlag":"OPS_FLAG_CLOSE",
-            "subTypeOfInferShape":"1",
-            "userDefined":"True",
-            "workspaceSize":"1024"
-        },
-        "output0":{
-            "name":"output",
-            "type":"DT_INT32,DT_INT64"
-        }
-    },
     "Mvlgamma":{
         "input0":{
             "name":"x",
@@ -7947,7 +7945,7 @@
     "ResizeBicubic":{
         "input0":{
             "name":"images",
-            "type":"DT_INT8,DT_UINT8,DT_INT16,DT_UINT16,DT_INT32,DT_INT64,DT_FLOAT16,DT_FLOAT,DT_DOUBLE"
+            "type":"DT_FLOAT16,DT_FLOAT,DT_DOUBLE"
         },
         "input1":{
             "name":"size",
@@ -7969,13 +7967,13 @@
         },
         "output0":{
             "name":"y",
-            "type":"DT_FLOAT"
+            "type":"DT_FLOAT16,DT_FLOAT,DT_DOUBLE"
         }
     },
     "ResizeBicubicGrad":{
         "input0":{
             "name":"grads",
-            "type":"DT_FLOAT"
+            "type":"DT_FLOAT,DT_DOUBLE"
         },
         "input1":{
             "name":"original_image",
@@ -8019,7 +8017,7 @@
             "kernelSo":"libcust_cpu_kernels.so",
             "opKernelLib":"CUSTAICPUKernel",
             "opsFlag":"OPS_FLAG_CLOSE",
-            "subTypeOfInferShape":"4",
+            "subTypeOfInferShape":"2",
             "userDefined":"True",
             "workspaceSize":"1024"
         },

mindspore/nn/cell.py

@@ -1081,9 +1081,6 @@ class Cell(Cell_):
         if not isinstance(param, Parameter) and param is not None:
             raise TypeError(f"For 'insert_param_to_cell', the argument 'param' must be 'Parameter' if not None, "
                             f"but got {type(param)}.")
-        if param is None:
-            raise TypeError(f"For 'insert_param_to_cell', the argument 'param' must not be None, "
-                            f"but got None.")
         if isinstance(param, Parameter) and param.name == PARAMETER_NAME_DEFAULT:
             param.name = param_name
         self._params[param_name] = param

mindspore/nn/layer/activation.py

@@ -932,10 +932,8 @@ class GELU(Cell):
         """Initialize GELU."""
         super(GELU, self).__init__()
         validator.check_bool(approximate, 'approximate', self.cls_name)
-        self.approximate = approximate
-        if approximate:
-            self.approximate = 'tanh'
-        else:
+        self.approximate = 'tanh'
+        if not approximate:
             self.approximate = 'none'
 
     def construct(self, x):
@@ -1335,7 +1333,8 @@ class LRN(Cell):
 
     .. warning::
         LRN is deprecated on Ascend due to potential accuracy problem. It's recommended to use other
-        normalization methods, e.g. :class:`mindspore.nn.BatchNorm`.
+        normalization methods, e.g. :class:`mindspore.nn.BatchNorm1d` ,
+        :class:`mindspore.nn.BatchNorm2d` , :class:`mindspore.nn.BatchNorm3d`.
 
     Refer to :func:`mindspore.ops.lrn` for more details.
 

mindspore/nn/layer/conv.py

@@ -718,9 +718,9 @@ class Conv3d(_Conv):
 
         .. math::
             \begin{array}{ll} \\
-                D_{out} ＝ \left \lceil{\frac{D_{in}}{\text{stride[0]}}} \right \rceil \\
-                H_{out} ＝ \left \lceil{\frac{H_{in}}{\text{stride[1]}}} \right \rceil \\
-                W_{out} ＝ \left \lceil{\frac{W_{in}}{\text{stride[2]}}} \right \rceil \\
+                D_{out} = \left \lceil{\frac{D_{in}}{\text{stride[0]}}} \right \rceil \\
+                H_{out} = \left \lceil{\frac{H_{in}}{\text{stride[1]}}} \right \rceil \\
+                W_{out} = \left \lceil{\frac{W_{in}}{\text{stride[2]}}} \right \rceil \\
             \end{array}
 
 
@@ -728,11 +728,11 @@ class Conv3d(_Conv):
 
         .. math::
             \begin{array}{ll} \\
-                D_{out} ＝ \left \lfloor{\frac{D_{in} - \text{dilation[0]} \times (\text{kernel_size[0]} - 1) }
+                D_{out} = \left \lfloor{\frac{D_{in} - \text{dilation[0]} \times (\text{kernel_size[0]} - 1) }
                 {\text{stride[0]}} + 1} \right \rfloor \\
-                H_{out} ＝ \left \lfloor{\frac{H_{in} - \text{dilation[1]} \times (\text{kernel_size[1]} - 1) }
+                H_{out} = \left \lfloor{\frac{H_{in} - \text{dilation[1]} \times (\text{kernel_size[1]} - 1) }
                 {\text{stride[1]}} + 1} \right \rfloor \\
-                W_{out} ＝ \left \lfloor{\frac{W_{in} - \text{dilation[2]} \times (\text{kernel_size[2]} - 1) }
+                W_{out} = \left \lfloor{\frac{W_{in} - \text{dilation[2]} \times (\text{kernel_size[2]} - 1) }
                 {\text{stride[2]}} + 1} \right \rfloor \\
             \end{array}
 
@@ -740,11 +740,11 @@ class Conv3d(_Conv):
 
         .. math::
             \begin{array}{ll} \\
-                D_{out} ＝ \left \lfloor{\frac{D_{in} + padding[0] + padding[1] - (\text{dilation[0]} - 1) \times
+                D_{out} = \left \lfloor{\frac{D_{in} + padding[0] + padding[1] - (\text{dilation[0]} - 1) \times
                 \text{kernel_size[0]} - 1 }{\text{stride[0]}} + 1} \right \rfloor \\
-                H_{out} ＝ \left \lfloor{\frac{H_{in} + padding[2] + padding[3] - (\text{dilation[1]} - 1) \times
+                H_{out} = \left \lfloor{\frac{H_{in} + padding[2] + padding[3] - (\text{dilation[1]} - 1) \times
                 \text{kernel_size[1]} - 1 }{\text{stride[1]}} + 1} \right \rfloor \\
-                W_{out} ＝ \left \lfloor{\frac{W_{in} + padding[4] + padding[5] - (\text{dilation[2]} - 1) \times
+                W_{out} = \left \lfloor{\frac{W_{in} + padding[4] + padding[5] - (\text{dilation[2]} - 1) \times
                 \text{kernel_size[2]} - 1 }{\text{stride[2]}} + 1} \right \rfloor \\
             \end{array}
 
@@ -812,7 +812,7 @@ class Conv3d(_Conv):
             bias_init,
             data_format,
             dtype=dtype)
-        out_channels = self.out_channels
+        out_channels = self.out_channels // group
         self.conv3d = P.Conv3D(out_channel=out_channels,
                                kernel_size=self.kernel_size,
                                mode=1,
@@ -820,17 +820,33 @@ class Conv3d(_Conv):
                                pad=self.padding,
                                stride=self.stride,
                                dilation=self.dilation,
-                               group=group,
+                               group=1,
                                data_format=self.data_format)
         self.bias_add = P.BiasAdd(data_format=self.data_format)
         self.shape = P.Shape()
+        self.concat = P.Concat(1)
+        self.split_0 = P.Split(0, self.group)
+        self.split_1 = P.Split(1, self.group)
 
     def construct(self, x):
         x_shape = self.shape(x)
         _check_input_5dims(x_shape, self.cls_name)
-        out = self.conv3d(x, self.weight)
-        if self.has_bias:
-            out = self.bias_add(out, self.bias)
+        if self.group == 1:
+            out = self.conv3d(x, self.weight)
+            if self.has_bias:
+                out = self.bias_add(out, self.bias)
+        else:
+            features = self.split_1(x)
+            weights = self.split_0(self.weight)
+            outputs = ()
+            for i in range(self.group):
+                output = self.conv3d(features[i], weights[i])
+                outputs = outputs + (output,)
+            out = self.concat(outputs)
+            if self.bias is not None:
+                new_shape = [1 for _ in range(out.ndim)]
+                new_shape[1] = self.out_channels
+                out = out + self.bias.reshape(new_shape)
         return out
 
 
@@ -921,9 +937,9 @@ class Conv3dTranspose(_Conv):
 
         .. math::
             \begin{array}{ll} \\
-                D_{out} ＝ \left \lfloor{\frac{D_{in}}{\text{stride[0]}} + 1} \right \rfloor \\
-                H_{out} ＝ \left \lfloor{\frac{H_{in}}{\text{stride[1]}} + 1} \right \rfloor \\
-                W_{out} ＝ \left \lfloor{\frac{W_{in}}{\text{stride[2]}} + 1} \right \rfloor \\
+                D_{out} = \left \lfloor{\frac{D_{in}}{\text{stride[0]}} + 1} \right \rfloor \\
+                H_{out} = \left \lfloor{\frac{H_{in}}{\text{stride[1]}} + 1} \right \rfloor \\
+                W_{out} = \left \lfloor{\frac{W_{in}}{\text{stride[2]}} + 1} \right \rfloor \\
             \end{array}
 
 
@@ -931,11 +947,11 @@ class Conv3dTranspose(_Conv):
 
         .. math::
             \begin{array}{ll} \\
-                D_{out} ＝ \left \lfloor{\frac{D_{in} - \text{dilation[0]} \times (\text{kernel_size[0]} - 1) }
+                D_{out} = \left \lfloor{\frac{D_{in} - \text{dilation[0]} \times (\text{kernel_size[0]} - 1) }
                 {\text{stride[0]}} + 1} \right \rfloor \\
-                H_{out} ＝ \left \lfloor{\frac{H_{in} - \text{dilation[1]} \times (\text{kernel_size[1]} - 1) }
+                H_{out} = \left \lfloor{\frac{H_{in} - \text{dilation[1]} \times (\text{kernel_size[1]} - 1) }
                 {\text{stride[1]}} + 1} \right \rfloor \\
-                W_{out} ＝ \left \lfloor{\frac{W_{in} - \text{dilation[2]} \times (\text{kernel_size[2]} - 1) }
+                W_{out} = \left \lfloor{\frac{W_{in} - \text{dilation[2]} \times (\text{kernel_size[2]} - 1) }
                 {\text{stride[2]}} + 1} \right \rfloor \\
             \end{array}
 
@@ -943,11 +959,11 @@ class Conv3dTranspose(_Conv):
 
         .. math::
             \begin{array}{ll} \\
-                D_{out} ＝ \left \lfloor{\frac{D_{in} + padding[0] + padding[1] - (\text{dilation[0]} - 1) \times
+                D_{out} = \left \lfloor{\frac{D_{in} + padding[0] + padding[1] - (\text{dilation[0]} - 1) \times
                 \text{kernel_size[0]} - 1 }{\text{stride[0]}} + 1} \right \rfloor \\
-                H_{out} ＝ \left \lfloor{\frac{H_{in} + padding[2] + padding[3] - (\text{dilation[1]} - 1) \times
+                H_{out} = \left \lfloor{\frac{H_{in} + padding[2] + padding[3] - (\text{dilation[1]} - 1) \times
                 \text{kernel_size[1]} - 1 }{\text{stride[1]}} + 1} \right \rfloor \\
-                W_{out} ＝ \left \lfloor{\frac{W_{in} + padding[4] + padding[5] - (\text{dilation[2]} - 1) \times
+                W_{out} = \left \lfloor{\frac{W_{in} + padding[4] + padding[5] - (\text{dilation[2]} - 1) \times
                 \text{kernel_size[2]} - 1 }{\text{stride[2]}} + 1} \right \rfloor \\
             \end{array}