!4859 Add CTCGrerdyDecoder ops for old backend.

Merge pull request !4859 from liuxiao93/Add-ReversSqueuce-EditDistance-CTCGrerdyDecoder

mindspore/ccsrc/transform/graph_ir/op_adapter_map.h

@@ -192,6 +192,8 @@ constexpr const char kNameAscendDequant[] = "Dequant";
 constexpr const char kNameReverseSequence[] = "ReverseSequence";
 constexpr const char kNameEditDistance[] = "EditDistance";
 constexpr const char kNameCase[] = "Case";
+constexpr const char kNameAssert[] = "Assert";
+constexpr const char kNameCTCGreedyDecoder[] = "CTCGreedyDecoder";
 
 class OpAdapterMap {
  public:

mindspore/ccsrc/transform/graph_ir/op_declare/ctc_ops_declare.cc

@@ -28,4 +28,13 @@ ATTR_MAP(CTCLoss) = {
   {"ignore_longer_outputs_than_inputs", ATTR_DESC(ignore_longer_outputs_than_inputs, AnyTraits<bool>())}};
 OUTPUT_MAP(CTCLoss) = {{0, OUTPUT_DESC(loss)}, {1, OUTPUT_DESC(gradient)}};
 REG_ADPT_DESC(CTCLoss, kNameCTCLoss, ADPT_DESC(CTCLoss))
+
+// CTCGreedyDecoder
+INPUT_MAP(CTCGreedyDecoder) = {{1, INPUT_DESC(inputs)}, {2, INPUT_DESC(sequence_length)}};
+ATTR_MAP(CTCGreedyDecoder) = {{"merge_repeated", ATTR_DESC(merge_repeated, AnyTraits<bool>())}};
+OUTPUT_MAP(CTCGreedyDecoder) = {{0, OUTPUT_DESC(decoded_indices)},
+                                {1, OUTPUT_DESC(decoded_values)},
+                                {2, OUTPUT_DESC(decoded_shape)},
+                                {3, OUTPUT_DESC(log_probability)}};
+REG_ADPT_DESC(CTCGreedyDecoder, kNameCTCGreedyDecoder, ADPT_DESC(CTCGreedyDecoder))
 }  // namespace mindspore::transform

mindspore/ccsrc/transform/graph_ir/op_declare/ctc_ops_declare.h

@@ -25,5 +25,8 @@
 namespace mindspore::transform {
 DECLARE_OP_ADAPTER(CTCLoss)
 DECLARE_OP_USE_OUTPUT(CTCLoss)
+
+DECLARE_OP_ADAPTER(CTCGreedyDecoder)
+DECLARE_OP_USE_OUTPUT(CTCGreedyDecoder)
 }  // namespace mindspore::transform
 #endif  // MINDSPORE_CCSRC_TRANSFORM_GRAPH_IR_OP_DECLARE_CTC_OPS_DECLARE_H_

mindspore/ccsrc/transform/graph_ir/op_declare/logging_ops_declare.cc

@@ -23,5 +23,10 @@ INPUT_MAP(Print) = EMPTY_INPUT_MAP;
 DYN_INPUT_MAP(Print) = {{1, DYN_INPUT_DESC(x)}};
 ATTR_MAP(Print) = EMPTY_ATTR_MAP;
 REG_ADPT_DESC(Print, kNamePrint, ADPT_DESC(Print))
+
+INPUT_MAP(Assert) = {{1, INPUT_DESC(input_condition)}};
+DYN_INPUT_MAP(Assert) = {{2, DYN_INPUT_DESC(input_data)}};
+ATTR_MAP(Assert) = {{"summarize", ATTR_DESC(summarize, AnyTraits<int>())}};
+REG_ADPT_DESC(Assert, kNameAssert, ADPT_DESC(Assert))
 #endif
 }  // namespace mindspore::transform

mindspore/ccsrc/transform/graph_ir/op_declare/logging_ops_declare.h

@@ -26,6 +26,9 @@ namespace mindspore::transform {
 #ifdef ENABLE_GE
 DECLARE_OP_ADAPTER(Print)
 DECLARE_OP_USE_DYN_INPUT(Print)
+
+DECLARE_OP_ADAPTER(Assert)
+DECLARE_OP_USE_DYN_INPUT(Assert)
 #endif
 }  // namespace mindspore::transform
 #endif  // MINDSPORE_CCSRC_TRANSFORM_GRAPH_IR_OP_DECLARE_LOGGING_OPS_DECLARE_H_

mindspore/ops/operations/__init__.py

@@ -39,7 +39,7 @@ from .comm_ops import (AllGather, AllReduce, _AlltoAll, ReduceScatter, Broadcast
                        _VirtualDiv, _GetTensorSlice,
                        _HostAllGather, _HostReduceScatter)
 from .debug_ops import (ImageSummary, InsertGradientOf, HookBackward, ScalarSummary,
-                        TensorSummary, HistogramSummary, Debug, Print)
+                        TensorSummary, HistogramSummary, Debug, Print, Assert)
 from .control_ops import ControlDepend, GeSwitch, Merge
 from .inner_ops import ScalarCast
 
@@ -64,7 +64,7 @@ from .nn_ops import (LSTM, SGD, Adam, FusedSparseAdam, FusedSparseLazyAdam, Appl
                      DropoutDoMask, DropoutGrad, Dropout,
                      DropoutGenMask, Flatten, FusedBatchNorm, FusedBatchNormEx, BNTrainingReduce, BNTrainingUpdate,
                      Gelu, Elu,
-                     GetNext, L2Normalize, LayerNorm, L2Loss, CTCLoss, CTCLossV2,
+                     GetNext, L2Normalize, LayerNorm, L2Loss, CTCLoss, CTCLossV2, CTCGreedyDecoder,
                      LogSoftmax,
                      MaxPool, DataFormatDimMap,
                      AvgPool, Conv2DBackpropInput, ConfusionMulGrad,
@@ -201,6 +201,7 @@ __all__ = [
     'HistogramSummary',
     "Debug",
     "Print",
+    "Assert",
     'InsertGradientOf',
     'HookBackward',
     'InvertPermutation',
@@ -225,6 +226,7 @@ __all__ = [
     'SmoothL1Loss',
     'L2Loss',
     'CTCLoss',
+    'CTCGreedyDecoder',
     'RNNTLoss',
     'ReduceAll',
     'ReduceAny',

mindspore/ops/operations/debug_ops.py

@@ -16,6 +16,7 @@
 """debug_ops"""
 from types import FunctionType, MethodType
 from ..._checkparam import Validator as validator
+from ..._checkparam import Rel
 from ...common import dtype as mstype
 from ..primitive import prim_attr_register, PrimitiveWithInfer, Primitive
 
@@ -364,3 +365,47 @@ class Debug(Primitive):
 
     def __call__(self, *args, **kwargs):
         pass
+
+
+class Assert(PrimitiveWithInfer):
+    """
+    Asserts that the given condition is true.
+    If input condition evaluates to false, print the list of tensor in data.
+
+    Args:
+        summarize (int): Print this many entries of each tensor.
+
+    Inputs:
+        - **condition** [Union[Tensor[bool], bool]] - The condition to evaluate.
+        - **input_data** (Union(tuple[Tensor], list[Tensor])) - The tensors to print out when condition is false.
+
+    Examples:
+        >>> class AssertDemo(nn.Cell):
+        >>>     def __init__(self):
+        >>>         super(AssertDemo, self).__init__()
+        >>>         self.assert = P.Assert(summarize=10)
+        >>>         self.add = P.TensorAdd()
+        >>>
+        >>>     def construct(self, x, y):
+        >>>         data = self.add(x, y)
+        >>>         self.assert(True, [data])
+        >>>         return data
+    """
+
+    @prim_attr_register
+    def __init__(self, summarize=3):
+        """init Assert"""
+        self.summarize = validator.check_value_type("summarize", summarize, [int], self.name)
+
+    def infer_shape(self, condition, inputs):
+        condition_len = len(condition)
+        validator.check_integer("condition's rank", condition_len, 1, Rel.LE, self.name)
+        if condition_len == 1:
+            validator.check_integer("condition[0]", condition[0], 1, Rel.EQ, self.name)
+        return [1]
+
+    def infer_dtype(self, condition, inputs):
+        validator.check_scalar_or_tensor_type_same({"condition": condition}, [mstype.bool_], self.name)
+        for dtype in inputs:
+            validator.check_subclass("input", dtype, [mstype.tensor], self.name)
+        return mstype.int32

mindspore/ops/operations/nn_ops.py

@@ -5173,6 +5173,7 @@ class CTCLoss(PrimitiveWithInfer):
         - **inputs** (Tensor) - The input Tensor should be a `3-D` tensor whose shape is
           :math:`(max_time, batch_size, num_classes)`. `num_classes` should be `num_labels + 1` classes, `num_labels`
           indicates the number of actual labels. Blank labels are reserved. Default blank label is `num_classes - 1`.
+          Data type must be float32 or float64.
         - **labels_indices** (Tensor) - The indices of labels. `labels_indices[i, :] == [b, t]` means `labels_values[i]`
           stores the id for `(batch b, time t)`. The type must be int64 and rank must be 2.
         - **labels_values** (Tensor) - A `1-D` input tensor. The values associated with the given batch and time. The
@@ -5222,10 +5223,6 @@ class CTCLoss(PrimitiveWithInfer):
         return batch_size, inputs
 
     def infer_dtype(self, inputs, labels_indices, labels_values, sequence_length):
-        validator.check_subclass("inputs_dtype", inputs, mstype.tensor, self.name)
-        validator.check_subclass("labels_indices_dtype", labels_indices, mstype.tensor, self.name)
-        validator.check_subclass("labels_values_dtype", labels_values, mstype.tensor, self.name)
-        validator.check_subclass("sequence_length_dtype", sequence_length, mstype.tensor, self.name)
         validator.check_tensor_type_same({"inputs_dtype": inputs}, [mstype.float32, mstype.double], self.name)
         validator.check_tensor_type_same({"labels_indices_dtype": labels_indices}, [mstype.int64], self.name)
         validator.check_tensor_type_same({"labels_values_dtype": labels_values}, [mstype.int32], self.name)
@@ -5233,6 +5230,72 @@ class CTCLoss(PrimitiveWithInfer):
         return inputs, inputs
 
 
+class CTCGreedyDecoder(PrimitiveWithInfer):
+    """
+    Performs greedy decoding on the logits given in inputs.
+
+    Args:
+        merge_repeated (bool): If True, merge repeated classes in output. Default: True.
+
+    Inputs:
+        - **inputs** (Tensor) - The input Tensor should be a `3-D` tensor whose shape is
+          :math:`(max_time, batch_size, num_classes)`. `num_classes` should be `num_labels + 1` classes, `num_labels`
+          indicates the number of actual labels. Blank labels are reserved. Default blank label is `num_classes - 1`.
+          Data type must be float32 or float64.
+        - **sequence_length** (Tensor) - A tensor containing sequence lengths with the shape of :math:`(batch_size)`.
+          The type must be int32. Each value in the tensor should not greater than `max_time`.
+
+    Outputs:
+        - **decoded_indices** (Tensor) - A tensor with shape of :math:`(total_decoded_outputs, 2)`.
+          Data type is int64.
+        - **decoded_values** (Tensor) - A tensor with shape of :math:`(total_decoded_outputs)`,
+          it stores the decoded classes. Data type is int64.
+        - **decoded_shape** (Tensor) - The value of tensor is :math:`[batch_size, max_decoded_legth]`.
+          Data type is int64.
+        - **log_probability** (Tensor) - A tensor with shape of :math:`(batch_size, 1)`,
+          containing sequence log-probability. Has the same type as `inputs`.
+
+    Examples:
+        >>>    class CTCGreedyDecoderNet(nn.Cell):
+        >>>        def __init__(self):
+        >>>            super(CTCGreedyDecoderNet, self).__init__()
+        >>>            self.ctc_greedy_decoder = P.CTCGreedyDecoder()
+        >>>            self.assert_op = P.Assert(300)
+        >>>
+        >>>        def construct(self, inputs, sequence_length):
+        >>>            out = self.ctc_greedy_decoder(inputs,sequence_length)
+        >>>            self.assert_op(True, (out[0], out[1], out[2], out[3]))
+        >>>            return out[2]
+        >>>
+        >>> inputs = Tensor(np.random.random((2, 2, 3)), mindspore.float32)
+        >>> sequence_length = Tensor(np.array([2, 2]), mindspore.int32)
+        >>> net = CTCGreedyDecoderNet()
+        >>> output = net(inputs, sequence_length)
+    """
+
+    @prim_attr_register
+    def __init__(self, merge_repeated=True):
+        self.merge_repeated = validator.check_value_type("merge_repeated", merge_repeated, [bool], self.name)
+
+    def infer_shape(self, inputs_shape, sequence_length_shape):
+        validator.check_integer("inputs rank", len(inputs_shape), 3, Rel.EQ, self.name)
+        validator.check_integer("sequence_length rank", len(sequence_length_shape), 1, Rel.EQ, self.name)
+        validator.check('inputs batch_size', inputs_shape[1], 'sequence_length batch_size',
+                        sequence_length_shape[0], Rel.EQ, self.name)
+        total_decoded_outputs = -1
+        decoded_indices_shape = [total_decoded_outputs, 2]
+        decoded_values = [total_decoded_outputs]
+        decoded_shape = [2]
+        log_probability_shape = [inputs_shape[1], 1]
+        return decoded_indices_shape, decoded_values, decoded_shape, log_probability_shape
+
+    def infer_dtype(self, inputs_dtype, sequence_length_dtype):
+        validator.check_tensor_type_same({"inputs_dtype": inputs_dtype}, [mstype.float32, mstype.double], self.name)
+        validator.check_tensor_type_same({"sequence_length_dtype": sequence_length_dtype}, [mstype.int32], self.name)
+        decoded_type = mstype.tensor_type(mstype.int64)
+        return decoded_type, decoded_type, decoded_type, inputs_dtype
+
+
 class BasicLSTMCell(PrimitiveWithInfer):
     r"""
     Performs the long short term memory(LSTM) on the input.
@@ -5361,6 +5424,7 @@ class InTopK(PrimitiveWithInfer):
         >>> result = in_top_k(x1, x2)
         [True  False]
     """
+
     @prim_attr_register
     def __init__(self, k):
         """Init InTopK"""

tests/ut/python/ops/test_ops.py

@@ -621,6 +621,18 @@ class UniformNet(nn.Cell):
         return out
 
 
+class CTCGreedyDecoderNet(nn.Cell):
+    def __init__(self):
+        super(CTCGreedyDecoderNet, self).__init__()
+        self.ctc_greedy_decoder = P.CTCGreedyDecoder()
+        self.assert_op = P.Assert(300)
+
+    def construct(self, inputs, sequence_length):
+        out = self.ctc_greedy_decoder(inputs,sequence_length)
+        self.assert_op(True, (out[0], out[1], out[2], out[3]))
+        return out[2]
+
+
 class StridedSliceNet(nn.Cell):
     def __init__(self):
         super(StridedSliceNet, self).__init__()
@@ -1672,6 +1684,10 @@ test_case_nn_ops = [
                         Tensor(np.array([1, 2, 3, 4]).astype(np.int32)),
                         Tensor(np.array([6, 6, 6, 6]).astype(np.int32))],
         'desc_bprop': [[4], [6, 4, 6]]}),
+    ('CTCGreedyDecoder', {
+        'block': CTCGreedyDecoderNet(),
+        'desc_inputs': [[2, 2, 3], Tensor(np.array([2, 2]).astype(np.int32))],
+        'skip': ['backward']}),
     ('L2Loss_1', {
         'block': P.L2Loss(),
         'desc_inputs': [Tensor(np.array([1, 2, 3, 4]), mstype.float32)],