/third_party/mindspore/mindspore/ops/operations/ |
D | comm_ops.py | 164 def infer_dtype(self, x_dtype): argument 165 validator.check_tensor_dtype_valid('x', x_dtype, target_dtypes, self.name) 166 return x_dtype 241 def infer_dtype(self, x_dtype): argument 242 validator.check_tensor_dtype_valid('x', x_dtype, target_dtypes, self.name) 243 return x_dtype 278 def infer_dtype(self, x_dtype, z_shape): argument 279 validator.check_tensor_dtype_valid('x', x_dtype, target_dtypes, self.name) 280 return x_dtype 309 def infer_dtype(self, x_dtype, z_dtype): argument [all …]
|
D | math_ops.py | 93 def do_infer_dtype(x_dtype, y_dtype, valid_dtype=mstype.number_type, prim_name=None): argument 95 args_type = {"x": x_dtype, "y": y_dtype} 97 if x_dtype in complex_types or y_dtype in complex_types: 106 if (x_dtype.element_type(), y_dtype.element_type()) not in tpye_infer_dict.keys(): 111 return tpye_infer_dict.get((x_dtype.element_type(), y_dtype.element_type())) 114 return x_dtype 116 def infer_dtype(self, x_dtype, y_dtype): argument 117 return _MathBinaryOp.do_infer_dtype(x_dtype, y_dtype, mstype.number_type, self.name) 1516 def infer_dtype(self, x_dtype): argument 1517 validator.check_tensor_dtype_valid("x", x_dtype, mstype.number_type, self.name) [all …]
|
D | _grad_ops.py | 131 def infer_dtype(self, x_dtype, dout_dtype): argument 132 args = {"x": x_dtype, "dout": dout_dtype} 134 return x_dtype 148 def infer_dtype(self, x_dtype, dout_dtype): argument 149 args = {"x": x_dtype, "dout": dout_dtype} 153 return x_dtype 171 def infer_dtype(self, x_dtype, dout_dtype): argument 172 args = {"x": x_dtype, "dout": dout_dtype} 175 return x_dtype 733 def infer_dtype(self, y_backprop_dtype, x_dtype, y_dtype): argument [all …]
|
D | array_ops.py | 69 def infer_dtype(self, x_dtype, indices_dtype, updates_dtype): argument 71 args = {"x": x_dtype, "updates": updates_dtype} 73 return x_dtype 109 def check_dtype(self, x_dtype, indices_dtype, updates_dtype): argument 111 args = {"x": x_dtype, "updates": updates_dtype} 686 def infer_dtype(self, x_dtype): argument 687 validator.check_subclass("x", x_dtype, mstype.tensor, self.name) 688 return x_dtype 1686 def infer_dtype(self, x_dtype): argument 1687 validator.check_subclass("input_x", x_dtype, mstype.tensor, self.name) [all …]
|
D | _inner_ops.py | 170 def infer_dtype(self, x_dtype): argument 171 validator.check_tensor_dtype_valid('x', x_dtype, [mstype.float32, mstype.int32], self.name) 172 return x_dtype 314 def infer_dtype(self, x_dtype, assist_dtype): argument 316 args = {"x": x_dtype, "assist": assist_dtype} 318 return x_dtype 363 def infer_dtype(self, x_dtype, assist_dtype): argument 365 args = {"x": x_dtype, "assist": assist_dtype} 367 return x_dtype 430 def infer_dtype(self, x_dtype): argument [all …]
|
D | nn_ops.py | 233 def infer_dtype(self, x_dtype): argument 234 …validator.check_tensor_dtype_valid("x_dtype", x_dtype, [mstype.float16, mstype.float32, mstype.flo… 236 return x_dtype 502 def infer_dtype(self, x_dtype): argument 503 … validator.check_tensor_dtype_valid('x', x_dtype, [mstype.float16, mstype.float32], self.name) 504 return x_dtype 556 def infer_dtype(self, x_dtype): argument 558 validator.check_tensor_dtype_valid('x', x_dtype, valid_dtypes, self.name) 559 return x_dtype 753 def infer_dtype(self, x_dtype): argument [all …]
|
D | _thor_ops.py | 528 def infer_dtype(self, x_dtype): argument 530 validator.check_tensor_dtype_valid('x', x_dtype, valid_dtypes, self.name) 531 return x_dtype 602 def infer_dtype(self, x_dtype): argument 605 validator.check_tensor_dtype_valid('x', x_dtype, valid_dtypes, self.name) 606 return x_dtype
|
/third_party/mindspore/mindspore/core/ops/ |
D | reluv2.cc | 30 …GetOutputMaskShape(const std::vector<int64_t> &input_shape, const std::shared_ptr<Type> &x_dtype) { in GetOutputMaskShape() argument 38 if (x_dtype == kUInt8 || x_dtype == kInt8) { in GetOutputMaskShape() 49 if (x_dtype == kUInt8 || x_dtype == kInt8) { in GetOutputMaskShape() 71 auto x_dtype = input_type->element(); in InferShape() local 72 auto mask_shape = GetOutputMaskShape(input_shape, x_dtype); in InferShape() 80 auto min_mask_shape = GetOutputMaskShape(min_shape, x_dtype); in InferShape() 81 auto max_mask_shape = GetOutputMaskShape(max_shape, x_dtype); in InferShape()
|
D | cos.cc | 32 auto x_dtype = input_args[0]->BuildType(); in CosInferType() local 33 (void)CheckAndConvertUtils::CheckTensorTypeValid("x", x_dtype, common_valid_types, prim->name()); in CosInferType() 34 return x_dtype; in CosInferType()
|
D | broadcast_to.cc | 69 auto x_dtype = input_args[0]->BuildType()->cast<TensorTypePtr>(); in BroadcastToInferType() local 71 (void)CheckAndConvertUtils::CheckSubClass("x_dtype", x_dtype, template_types, prim->name()); in BroadcastToInferType() 72 return x_dtype->element(); in BroadcastToInferType()
|
D | tile.cc | 87 auto x_dtype = x_type_map->cast<TensorTypePtr>(); in TileInferType() local 88 MS_EXCEPTION_IF_NULL(x_dtype); in TileInferType() 90 return CheckAndConvertUtils::CheckTensorTypeValid("x_dtype", x_dtype, template_types, prim_name); in TileInferType()
|
D | square.cc | 45 auto x_dtype = input_args[kInputIndex0]->BuildType(); in SquareInferType() local 46 (void)CheckAndConvertUtils::CheckTensorTypeValid("x", x_dtype, common_valid_types, prim->name()); in SquareInferType() 47 return x_dtype; in SquareInferType()
|
D | diag.cc | 39 auto x_dtype = input_args[0]->BuildType(); in PartInferType() local 40 …return CheckAndConvertUtils::CheckTensorTypeValid("input type", x_dtype, common_valid_types, primi… in PartInferType()
|
D | diag_part.cc | 49 auto x_dtype = input_args[0]->BuildType(); in DiagPartInferType() local 50 …return CheckAndConvertUtils::CheckTensorTypeValid("input type", x_dtype, common_valid_types, primi… in DiagPartInferType()
|
D | arg_min.cc | 60 auto x_dtype = input_args[0]->BuildType()->cast<TensorTypePtr>()->element(); in ArgMinInfer() local 61 …return std::make_shared<abstract::AbstractTensor>(x_dtype, std::make_shared<abstract::Shape>(out_s… in ArgMinInfer()
|
/third_party/mindspore/mindspore/ops/_op_impl/_custom_op/ |
D | minmax_update_perchannel.py | 85 x_dtype = x.get("dtype") 104 x_dtype = x_dtype.lower() 107 util.check_dtype_rule(x_dtype, check_list) 115 input_data = tvm.placeholder(x.get("shape"), name="x", dtype=x_dtype) 116 min_data = tvm.placeholder(shape_c, name="min_val", dtype=x_dtype) 117 max_data = tvm.placeholder(shape_c, name="max_val", dtype=x_dtype)
|
D | fake_quant_perchannel.py | 95 x_dtype = x.get("dtype") 114 x_dtype = x_dtype.lower() 117 util.check_dtype_rule(x_dtype, check_list) 125 return x_shape, shape_c, x_dtype 138 x_shape, shape_c, x_dtype = fake_quant_perchannel_param(x, min_val, max_val, 140 input_data = tvm.placeholder(x_shape, name="x", dtype=x_dtype) 141 min_data = tvm.placeholder(shape_c, name="min_val", dtype=x_dtype) 142 max_data = tvm.placeholder(shape_c, name="max_val", dtype=x_dtype)
|
D | fake_quant_perchannel_grad.py | 119 x_dtype = x.get("dtype") 138 x_dtype = x_dtype.lower() 141 util.check_dtype_rule(x_dtype, check_list) 149 return x_shape, shape_c, x_dtype 166 x_shape, shape_c, x_dtype = fake_quant_perchannel_grad_param(x, min_val, max_val, 168 dout_data = tvm.placeholder(x_shape, name="dout", dtype=x_dtype) 169 input_data = tvm.placeholder(x_shape, name="x", dtype=x_dtype) 170 min_data = tvm.placeholder(shape_c, name="min_val", dtype=x_dtype) 171 max_data = tvm.placeholder(shape_c, name="max_val", dtype=x_dtype)
|
D | minmax_update_perlayer.py | 98 x_dtype = input_dtype.lower() 101 util.check_dtype_rule(x_dtype, check_list) 108 input_data = tvm.placeholder(input_shape, name="x", dtype=x_dtype)
|
D | fake_quant_perlayer_grad.py | 138 x_dtype = input_dtype.lower() 141 util.check_dtype_rule(x_dtype, check_list) 153 dout_data = tvm.placeholder(input_shape, name="dout", dtype=x_dtype) 154 input_data = tvm.placeholder(input_shape, name="x", dtype=x_dtype)
|
D | fake_quant_perlayer.py | 112 x_dtype = input_dtype.lower() 115 util.check_dtype_rule(x_dtype, check_list) 127 input_data = tvm.placeholder(input_shape, name="x", dtype=x_dtype)
|
/third_party/mindspore/mindspore/ccsrc/backend/optimizer/ascend/mindir/ |
D | avg_pool_grad_unify_mindir.cc | 102 const PadMode pad_mode, const TypeId x_dtype) { in CreateMeanMatrixValueNode() argument 146 …auto output_tensor = std::make_shared<tensor::Tensor>(x_dtype, output_shape, &output[0], kNumberTy… in CreateMeanMatrixValueNode() 148 auto abstract = std::make_shared<abstract::AbstractTensor>(TypeIdToType(x_dtype), output_shape); in CreateMeanMatrixValueNode() 157 const std::vector<int64_t> &k_size, const TypeId x_dtype) { in CreateKernelMatrixValueNode() argument 168 …auto kernel_matrix_tensor = std::make_shared<tensor::Tensor>(x_dtype, kernel_shape, &data[0], kNum… in CreateKernelMatrixValueNode() 170 auto abstract = std::make_shared<abstract::AbstractTensor>(TypeIdToType(x_dtype), kernel_shape); in CreateKernelMatrixValueNode() 194 auto x_dtype = AnfAlgo::GetPrevNodeOutputInferDataType(avgpool_grad, 0); in Process() local 200 …o mean_matrix_vnode = CreateMeanMatrixValueNode(graph, x_shape, k_size, stride, pad_mode, x_dtype); in Process() 201 auto kernel_matrix_vnode = CreateKernelMatrixValueNode(graph, x_shape, k_size, x_dtype); in Process()
|
/third_party/mindspore/mindspore/nn/layer/ |
D | basic.py | 351 def _dtype_check(x_dtype, prim_name=None): argument 353 if x_dtype not in [mstype.float32, mstype.float16]: 999 def tril(x_shape, x_dtype, k): argument 1003 return Tensor(mask, x_dtype) 1086 def triu(x_shape, x_dtype, k): argument 1090 return Tensor(mask, x_dtype) 1173 def _get_matrix_diag_assist(x_shape, x_dtype): argument 1177 return Tensor(assist, x_dtype) 1181 def _get_matrix_diag_part_assist(x_shape, x_dtype): argument 1185 return Tensor(assist, x_dtype) [all …]
|
/third_party/mindspore/mindspore/ops/_grad/ |
D | grad_inner_ops.py | 26 def _get_matrix_diag_assist(x_shape, x_dtype): argument 27 base_eye = P.Eye()(x_shape[-1], x_shape[-1], x_dtype).flatten() 33 def _get_matrix_diag_part_assist(x_shape, x_dtype): argument 34 base_eye = P.Eye()(x_shape[-2], x_shape[-1], x_dtype).flatten()
|
/third_party/mindspore/mindspore/ccsrc/backend/optimizer/ascend/ir_fission/ |
D | space_to_depth_split.cc | 108 TypeId x_dtype = AnfAlgo::GetOutputInferDataType(ori_inputs[kIndex1], 0); in Process() local 109 if (x_dtype != kNumberTypeFloat16) { in Process() 110 …INFO) << "Node " << cnode->DebugString() << ": The data type of node's first input is: " << x_dtype in Process()
|