[Lod] [fluid_ops] LodTensor python (#69484)

co63oc · web-flow · commit 42e3d60e03a2 · 2024-11-20T11:46:17.000+08:00
diff --git a/python/paddle/autograd/saved_tensors_hooks.py b/python/paddle/autograd/saved_tensors_hooks.py
@@ -36,7 +36,7 @@ class saved_tensors_hooks:
             of the original tensor. `pack_hook` will also be called while any
             tensor need be saved by `PyLayerContext.save_for_backward`. If a tensor
             saved for backward is no need buffer, `pack_hook` will not be called.
-            Only the tensor saved for backward is LoDTensor, `pack_hook` will be
+            Only the tensor saved for backward is DenseTensor, `pack_hook` will be
             called.
         unpack_hook (function): The unpack hook will be called every time the
             backward need use the saved inputs/outputs tensors. Then you can reload
diff --git a/python/paddle/base/data_feed_desc.py b/python/paddle/base/data_feed_desc.py
@@ -131,7 +131,7 @@ def set_dense_slots(self, dense_slots_name):
         Set slots in :attr:`dense_slots_name` as dense slots. **Note: In default, all slots are sparse slots.**
 
         Features for a dense slot will be fed into a Tensor, while those for a
-        sparse slot will be fed into a LoDTensor.
+        sparse slot will be fed into a DenseTensor.
 
         Examples:
             .. code-block:: python
diff --git a/python/paddle/base/executor.py b/python/paddle/base/executor.py
@@ -257,7 +257,7 @@ def check_feed_shape_type(var, feed, num_places=1):
 
     Args:
         var (Variable): the Variable object
-        feed (LoDTensor): the fed value, which must be a LoDTensor
+        feed (DenseTensor): the fed value, which must be a DenseTensor
         num_places: an integer value indicating the number of places.
             ParallelExecutor will divide data into devices (CPU/GPU) evenly.
     Returns:
@@ -303,7 +303,7 @@ def pir_check_feed_shape_type(feed, name, target_shape, dtype, num_places=1):
        is compatible with any number.
 
     Args:
-        feed (LoDTensor): the fed value, which must be a LoDTensor
+        feed (DenseTensor): the fed value, which must be a DenseTensor
         name (str): name of the variable
         target_shape (list): the shape that will be compared with feed
         dtype (core.VarDesc.VarType): the dtype that will be compared with feed
@@ -395,7 +395,7 @@ def has_fetch_operators(
         fetch_targets: a dictionary of {fetch_target_name: fetch_target_data}
         fetch_holder_name: the name of the variable that holds the data of
             all fetch targets. The type of this fetch_holder variable is
-            FETCH_LIST, which is essentially vector<LoDTensor>.
+            FETCH_LIST, which is essentially vector<DenseTensor>.
         fetch_op: the operator name of fetch
 
     Return:
@@ -601,7 +601,7 @@ def _fetch_var(name, scope=None, return_numpy=True):
             Default True.
 
     Returns:
-       LodTensor|numpy.ndarray
+       DenseTensor|numpy.ndarray
     """
     assert isinstance(name, str)
     if scope is None:
@@ -737,7 +737,7 @@ def _as_lodtensor(data, place, dtype=None):
         dtype(core.VarDesc.VarType|str): the expected data type of created tensor
 
     Returns:
-        LoDTensor
+        DenseTensor
     """
     # NOTE(zhiqiu): convert python builtin, like float, int, and list, to numpy ndarray
     if not isinstance(data, np.ndarray):
@@ -840,7 +840,7 @@ def run(
                 after the model runs. The default is None.
             return_numpy(bool): This parameter indicates whether convert the fetched Tensors
                 (the Tensor specified in the fetch list) to numpy.ndarray. if it is False,
-                the type of the return value is a list of :code:`LoDTensor`. The default is True.
+                the type of the return value is a list of :code:`DenseTensor`. The default is True.
         """
         tensors = self._new_exe.run(
             feed_names, enable_job_schedule_profiler
@@ -1777,7 +1777,7 @@ def run(
                 it to different scope. default is :code:`paddle.static.global_scope()`
             return_numpy(bool): This parameter indicates whether convert the fetched Tensors
                 (the Tensor specified in the fetch list) to numpy.ndarray. if it is False,
-                the type of the return value is a list of :code:`LoDTensor`. The default is True.
+                the type of the return value is a list of :code:`DenseTensor`. The default is True.
             use_program_cache(bool): This parameter indicates whether the input :code:`Program` is cached.
                 If the parameter is True, the model may run faster in the following cases:
                 the input program is :code:`paddle.static.Program`, and the parameters(program, feed Tensor name
@@ -1800,6 +1800,7 @@ def run(
             .. code-block:: python
                 :name: code-example-1
 
+                >>> # doctest: +SKIP("This has diff in xdoctest env")
                 >>> import paddle
                 >>> import numpy
 
@@ -1832,6 +1833,7 @@ def run(
             .. code-block:: python
                 :name: code-example-2
 
+                >>> # doctest: +SKIP("This has diff in xdoctest env")
                 >>> # doctest: +REQUIRES(env:GPU)
                 >>> import paddle
                 >>> import numpy as np
diff --git a/python/paddle/base/framework.py b/python/paddle/base/framework.py
@@ -1946,7 +1946,7 @@ def gradient(self):
         Get the Gradient of Current Variable
 
         Returns:
-            ndarray or tuple of ndarray: if Variable's type is LoDTensor, return numpy value of the gradient of current Variable, if Variable's type is SelectedRows, return tuple of ndarray, first element of tuple is numpy value of the gradient of current Variable, second element of tuple is numpy value of the rows of current Variable.
+            ndarray or tuple of ndarray: if Variable's type is DenseTensor, return numpy value of the gradient of current Variable, if Variable's type is SelectedRows, return tuple of ndarray, first element of tuple is numpy value of the gradient of current Variable, second element of tuple is numpy value of the rows of current Variable.
 
         Examples:
             .. code-block:: python
diff --git a/python/paddle/base/lod_tensor.py b/python/paddle/base/lod_tensor.py
@@ -22,7 +22,7 @@
 
 def create_lod_tensor(data, recursive_seq_lens, place):
     """
-    Create a DenseTensor from a numpy array, list or existing LoDTensor.
+    Create a DenseTensor from a numpy array, list or existing DenseTensor.
 
     The implementation is as follows:
 
@@ -34,7 +34,7 @@ def create_lod_tensor(data, recursive_seq_lens, place):
     3. Based on :code:`place` , copy the :code:`data` from a numpy array, list
        or existing DenseTensor to CPU or GPU device.
 
-    4. Set offset-based LoD to the output LoDTensor.
+    4. Set offset-based LoD to the output DenseTensor.
 
     Suppose we want to create a DenseTensor to hold data for word sequences,
     where each word is represented by an integer. If we want to create
@@ -48,7 +48,7 @@ def create_lod_tensor(data, recursive_seq_lens, place):
 
 
     Args:
-        data (numpy.ndarray|list|LoDTensor): a numpy array, a list or ad LoDTensor
+        data (numpy.ndarray|list|DenseTensor): a numpy array, a list or ad DenseTensor
                 holding the data to be copied.
         recursive_seq_lens (list[list[int]]): a list of lists indicating the
                 length-based LoD info.
@@ -126,7 +126,7 @@ def create_random_int_lodtensor(
 
     2. Create a numpy array of random integers, and parse the created numpy
        array as parameter :code:`data` of :ref:`api_paddle_base_create_lod_tensor` to
-       create the output LoDTensor.
+       create the output DenseTensor.
 
     Suppose we want to create a DenseTensor to hold data for 2 sequences, where
     the dimension of the sequences are [2, 30] and [3, 30] respectively.
diff --git a/python/paddle/base/reader.py b/python/paddle/base/reader.py
@@ -207,9 +207,9 @@ def from_generator(
             return_list (bool, optional): whether the return value on each device is
                 presented as a list. It is only valid when iterable=True.
                 If return_list=False, the return value on each device would
-                be a dict of str -> LoDTensor, where the key of the dict is
+                be a dict of str -> DenseTensor, where the key of the dict is
                 the name of each fed Tensors. If return_list=True, the
-                return value on each device would be a list(LoDTensor). It is
+                return value on each device would be a list(DenseTensor). It is
                 recommended to use return_list=False in static graph mode and
                 use return_list=True in dygraph mode.
             use_multiprocess (bool, optional): whether to use multi-process to
@@ -1153,9 +1153,9 @@ class PyReader(DataLoaderBase):
         return_list (bool): whether the return value on each device is
             presented as a list. It is only valid when iterable=True.
             If return_list=False, the return value on each device would
-            be a dict of str -> LoDTensor, where the key of the dict is
+            be a dict of str -> DenseTensor, where the key of the dict is
             the name of each fed variables. If return_list=True, the
-            return value on each device would be a list(LoDTensor). It is
+            return value on each device would be a list(DenseTensor). It is
             recommended to use return_list=False in static graph mode and
             use return_list=True in dygraph mode.
 
@@ -1557,12 +1557,12 @@ def decorate_batch_generator(self, reader, places=None):
         Set the data source of the PyReader object.
 
         The provided :code:`reader` should be a Python generator,
-        which yields numpy.ndarray-typed or LoDTensor-typed batched data.
+        which yields numpy.ndarray-typed or DenseTensor-typed batched data.
 
         :code:`places` must be set when the PyReader object is iterable.
 
         Args:
-            reader (generator): Python generator that yields LoDTensor-typed
+            reader (generator): Python generator that yields DenseTensor-typed
                 batched data.
             places (None|list(CUDAPlace)|list(CPUPlace)): place list. Must
                 be provided when PyReader is iterable.
diff --git a/python/paddle/framework/io.py b/python/paddle/framework/io.py
@@ -547,7 +547,7 @@ def condition(obj):
 
 def _transformed_from_varbase(obj):
     # In paddle2.1 version, Tensor is saved as tuple(tensor.name, tensor.numpy()).
-    # When executing paddle.load, use this function to determine whether to restore to Tensor/LoDTensor.
+    # When executing paddle.load, use this function to determine whether to restore to Tensor.
     if isinstance(obj, tuple) and len(obj) == 2:
         name_types = str
         if isinstance(obj[0], name_types) and isinstance(obj[1], np.ndarray):
@@ -557,7 +557,7 @@ def _transformed_from_varbase(obj):
 
 def _transformed_from_lodtensor(obj):
     # In paddle2.1 version, DenseTensor is saved as np.array(tensor).
-    # When executing paddle.load, use this function to determine whether to restore to Tensor/LoDTensor.
+    # When executing paddle.load, use this function to determine whether to restore to Tensor.
     if isinstance(obj, np.ndarray):
         return True
     return False
@@ -665,7 +665,7 @@ def ndarray_to_tensor(obj):
             obj, _transformed_from_varbase, tuple_to_tensor
         )
     # If there is no tuple(name, ndarray), it is considered to be saved by paddle2.0
-    # or converted from LoDTensor, and all ndarrays are converted to tensor.
+    # or converted from DenseTensor, and all ndarrays are converted to tensor.
     else:
         return _parse_every_object(
             obj, _transformed_from_lodtensor, ndarray_to_tensor
@@ -764,7 +764,7 @@ def _save_binary_var(obj, path):
     elif isinstance(obj, core.eager.Tensor):
         _save_lod_tensor(obj.value().get_tensor(), path)
     else:
-        # Since the concept of 'Tensor' is only exposed to users, the error message can only contain tensor instead of 'LoDTensor' or 'SelectedRows'
+        # Since the concept of 'Tensor' is only exposed to users, the error message can only contain tensor instead of 'DenseTensor' or 'SelectedRows'
         raise NotImplementedError(
             f"When use_binary_format = True, `paddle.save`  expected Tensor, but received {type(obj)}."
         )
diff --git a/python/paddle/incubate/layers/nn.py b/python/paddle/incubate/layers/nn.py
@@ -62,7 +62,7 @@ def fused_seqpool_cvm(
     **Note:** The Op only receives List of DenseTensor as input, only support SUM pooling now.
 
     Args:
-        input(Tensor): Input is List of LoDTensor.
+        input(Tensor): Input is List of DenseTensor.
         pool_type(str): pooling type, only support SUM pooling now.
         cvm(Tensor): cvm Tensor.
         pad_value(float, optional): padding value of sequence pool. Default: 0.0.
@@ -153,7 +153,7 @@ def search_pyramid_hash(
     **Pyramid hash embedding**
 
     Args:
-        input (Tensor): LoDTensor<int32> Tensor contained the IDs' information.
+        input (Tensor): DenseTensor<int32> Tensor contained the IDs' information.
         num_emb (int): The embedding size of output.
         space_len (int): The length of pyramid hash embedding space.
         pyramid_layer (int): The number of pyramid layers. It should be greater than 2.
@@ -273,7 +273,7 @@ def search_pyramid_hash(
 
 def shuffle_batch(x: Tensor, seed: int | Tensor | None = None) -> Tensor:
     """
-    This layer shuffle input tensor :attr:`x` . Normally, :attr:`x` is 2-D LoDTensor.
+    This layer shuffle input tensor :attr:`x` . Normally, :attr:`x` is 2-D DenseTensor.
 
     :attr:`x` is a DenseTensor to be shuffled with shape :math:`[N_1, N_2, ..., N_k, D]` . Note that the last dim of input will not be shuffled.
     :math:`N_1 * N_2 * ... * N_k` numbers of elements with length :math:`D` will be shuffled randomly.
@@ -350,7 +350,7 @@ def partial_concat(
     **Partial Concat**
     This OP concatenates the inputs according to the start index and length. This
     OP exists in incubate layers, which means that it is not shown to the public.
-    Only 2-D Tensor or LodTensor input is supported. Slice and concat can only be
+    Only 2-D Tensor input is supported. Slice and concat can only be
     performed along the second dimension.
 
     .. code-block:: text
@@ -430,7 +430,7 @@ def partial_sum(
     **PartialSum**
     This Op can sum the vars by specifying the initial position(start_index) and length(length).
     This Op exists in incubate layers, which means that it is not shown to the public.
-    Only 2-D Tensor or LodTensor input is supported. Slice and concat can only be
+    Only 2-D Tensor input is supported. Slice and concat can only be
     performed along the second dimension.
 
     .. code-block:: text
diff --git a/python/paddle/jit/api.py b/python/paddle/jit/api.py
@@ -1536,7 +1536,7 @@ def load(
     .. note::
         If you load model saved by ``paddle.static.save_inference_model`` ,
         there will be the following limitations when using it in fine-tuning:
-        1. Imperative mode do not support LoDTensor. All original model's feed targets or parameters that depend on LoD are temporarily unavailable.
+        1. Imperative mode do not support DenseTensor. All original model's feed targets or parameters that depend on LoD are temporarily unavailable.
         2. All saved model's feed targets need to be passed into TranslatedLayer's forward function.
         3. The variable's ``stop_gradient`` information is lost and can not be recovered.
         4. The parameter's ``trainable`` information is lost and can not be recovered.
diff --git a/python/paddle/jit/dy2static/partial_program.py b/python/paddle/jit/dy2static/partial_program.py
@@ -1074,8 +1074,8 @@ def _remove_no_value(self, out_vars):
 
     def _set_grad_type(self, params, train_program):
         # NOTE: if user set sparse gradient mode, the param's gradient
-        # will be SelectedRows, not LoDTensor. But tracer will just
-        # set param grad Tensor by forward Tensor(LoDTensor)
+        # will be SelectedRows, not DenseTensor. But tracer will just
+        # set param grad Tensor by forward Tensor(DenseTensor)
         # If we don't change grad_var type here, RunProgramOp need
         # transform SelectedRows to DenseTensor forcibly, it may not
         # be user wanted result.
diff --git a/python/paddle/jit/dy2static/pir_partial_program.py b/python/paddle/jit/dy2static/pir_partial_program.py
@@ -1211,8 +1211,8 @@ def _remove_no_value(self, out_vars):
 
     def _set_grad_type(self, params, train_program: RunnableProgram):
         # NOTE: if user set sparse gradient mode, the param's gradient
-        # will be SelectedRows, not LoDTensor. But tracer will just
-        # set param grad Tensor by forward Tensor(LoDTensor)
+        # will be SelectedRows, not DenseTensor. But tracer will just
+        # set param grad Tensor by forward Tensor(DenseTensor)
         # If we don't change grad_var type here, RunProgramOp need
         # transform SelectedRows to DenseTensor forcibly, it may not
         # be user wanted result.
diff --git a/python/paddle/jit/dy2static/utils.py b/python/paddle/jit/dy2static/utils.py
@@ -93,7 +93,7 @@ def data_layer_not_check(name, shape, dtype='float32', lod_level=0):
        dtype (np.dtype|VarType|str, optional): The type of the data. Supported
            dtype: bool, float16, float32, float64, int8, int16, int32, int64,
            uint8. Default: float32
-       lod_level (int, optional): The LoD level of the LoDTensor. Usually users
+       lod_level (int, optional): The LoD level of the DenseTensor. Usually users
            don't have to set this value. Default: 0
 
     Returns:
diff --git a/python/paddle/jit/translated_layer.py b/python/paddle/jit/translated_layer.py
@@ -1008,8 +1008,8 @@ def _run_dygraph(instance, input, program_holder):
 
     # NOTE: [ why need set param's gradient type here ]
     # if user set sparse gradient mode, the param's gradient
-    # will be SelectedRows, not LoDTensor. But tracer will just
-    # set param grad Tensor by forward Tensor(LoDTensor)
+    # will be SelectedRows, not DenseTensor. But tracer will just
+    # set param grad Tensor by forward Tensor(DenseTensor)
     # If we don't change grad_var type here, RunProgramOp need
     # transform SelectedRows to DenseTensor forcibly, it may not
     # be user wanted result.
diff --git a/python/paddle/nn/clip.py b/python/paddle/nn/clip.py
@@ -158,7 +158,7 @@ def get_tensor_from_selected_rows(x, name=None):
            x.height = 20
            x.value = [[1, 1] [2, 2] [2, 2] [3, 3] [6, 6]]
 
-        Output is LoDTensor:
+        Output is DenseTensor:
            out.shape = [5, 2]
            out.data = [[1, 1],
                        [2, 2],
diff --git a/python/paddle/nn/functional/extension.py b/python/paddle/nn/functional/extension.py
@@ -88,7 +88,7 @@ def sequence_mask(
     Args:
         x (Variable): Input tensor of sequence_mask layer, \
             whose elements are integers less than :code:`maxlen`. \
-            Tensor or LodTensor with shape [d_1, d_2, ..., d_n].
+            Tensor with shape [d_1, d_2, ..., d_n].
         maxlen (int|None, optional): Maximum length of the sequence. If :code:`maxlen` \
                            is None, it would be replace with :math:`max(x)`.
         dtype (np.dtype|paddle.dtype|str, optional): Data type of the output, \
diff --git a/python/paddle/static/input.py b/python/paddle/static/input.py
@@ -80,7 +80,7 @@ def data(
             dtype: bool, float16, float32, float64, int8, int16, int32, int64,
             uint8. Default: None. When `dtype` is not set, the dtype will get
             from the global dtype by `paddle.get_default_dtype()`.
-        lod_level (int, optional): The LoD level of the LoDTensor. Usually users
+        lod_level (int, optional): The LoD level of the DenseTensor. Usually users
             don't have to set this value. Default: 0.
 
     Returns:
@@ -89,6 +89,7 @@ def data(
     Examples:
         .. code-block:: python
 
+            >>> # doctest: +SKIP("This has diff in xdoctest env")
             >>> import numpy as np
             >>> import paddle
             >>> paddle.enable_static()
diff --git a/python/paddle/static/nn/common.py b/python/paddle/static/nn/common.py
diff --git a/python/paddle/tensor/array.py b/python/paddle/tensor/array.py
diff --git a/python/paddle/tensor/linalg.py b/python/paddle/tensor/linalg.py
diff --git a/python/paddle/tensor/manipulation.py b/python/paddle/tensor/manipulation.py
diff --git a/python/paddle/tensor/tensor.prototype.pyi b/python/paddle/tensor/tensor.prototype.pyi
