/aosp_15_r20/external/tensorflow/tensorflow/core/api_def/base_api/ |
H A D | api_def_TensorArrayGradV3.pbtxt | 9 The handle to the forward TensorArray. 21 The gradient source string, used to decide which gradient TensorArray 25 summary: "Creates a TensorArray for storing the gradients of values in the given handle." 27 If the given TensorArray gradient already exists, returns a reference to it. 29 Locks the size of the original TensorArray by disabling its dynamic size flag. 35 the forward TensorArray is dynamically sized, writes to this TensorArray 36 may resize the object. The gradient TensorArray is statically sized based 37 on the size of the forward TensorArray when this operation executes. 38 Furthermore, the size of the forward TensorArray is frozen by this call. 40 TensorArray only happens after all writes are executed. [all …]
|
H A D | api_def_TensorArrayV3.pbtxt | 4 name: "TensorArray" 15 The handle to the TensorArray. 34 validate the shapes of TensorArray elements. If this shape is not 41 A boolean that determines whether writes to the TensorArray 48 If true (default), Tensors in the TensorArray are cleared 57 elements in the TensorArray will be expected to have identical shapes. 68 resource. Default value is the name of the 'TensorArray' op (which
|
H A D | api_def_TensorArrayGradWithShape.pbtxt | 9 The handle to the forward TensorArray. 23 elements in the TensorArray corresponding to the input handle. 29 The gradient source string, used to decide which gradient TensorArray 33 summary: "Creates a TensorArray for storing multiple gradients of values in the given handle." 36 expanded shape compared to the input TensorArray whose gradient is being 37 computed. This enables multiple gradients for the same TensorArray to be
|
H A D | api_def_TensorArraySplitV3.pbtxt | 9 The handle to a TensorArray. 15 The concatenated tensor to write to the TensorArray. 22 TensorArray. 37 summary: "Split the data from the input value into TensorArray elements." 47 this splits values into a TensorArray with T tensors. 49 TensorArray index t will be the subtensor of values with starting position
|
H A D | api_def_TensorArrayGatherV3.pbtxt | 9 The handle to a TensorArray. 15 The locations in the TensorArray from which to read tensor elements. 27 All of the elements in the TensorArray, concatenated along a new 41 validate the shapes of TensorArray elements. If this shape is not 45 summary: "Gather specific elements from the TensorArray into output `value`."
|
H A D | api_def_TensorArraySizeV3.pbtxt | 9 The handle to a TensorArray (output of TensorArray or TensorArrayGrad). 21 The current size of the TensorArray. 24 summary: "Get the current size of the TensorArray."
|
H A D | api_def_TensorArrayConcatV3.pbtxt | 9 The handle to a TensorArray. 21 All of the elements in the TensorArray, concatenated along the first 44 TensorArray elements. If this shape is not fully specified, concatenating 48 summary: "Concat the elements from the TensorArray into value `value`."
|
H A D | api_def_TensorArrayCloseV3.pbtxt | 9 The handle to a TensorArray (output of TensorArray or TensorArrayGrad). 12 summary: "Delete the TensorArray from its resource container."
|
H A D | api_def_TensorArrayReadV3.pbtxt | 9 The handle to a TensorArray. 21 The tensor that is read from the TensorArray. 30 summary: "Read an element from the TensorArray into output `value`."
|
H A D | api_def_TensorArrayWriteV3.pbtxt | 9 The handle to a TensorArray. 15 The position to write to inside the TensorArray. 21 The tensor to write to the TensorArray.
|
H A D | api_def_TensorArrayScatterV3.pbtxt | 9 The handle to a TensorArray. 21 The concatenated tensor to write to the TensorArray. 36 summary: "Scatter the data from the input value into specific TensorArray elements."
|
/aosp_15_r20/external/tensorflow/tensorflow/compiler/tests/ |
H A D | tensor_array_ops_test.py | 56 ta = tensor_array_ops.TensorArray( 80 ta = tensor_array_ops.TensorArray( 101 ta = tensor_array_ops.TensorArray( 118 ta = tensor_array_ops.TensorArray( 142 ta = tensor_array_ops.TensorArray( 159 ta = tensor_array_ops.TensorArray( 179 ta = tensor_array_ops.TensorArray( 206 ta = tensor_array_ops.TensorArray( 225 ta = tensor_array_ops.TensorArray( 242 ta = tensor_array_ops.TensorArray( [all …]
|
H A D | cond_test.py | 45 ta = tensor_array_ops.TensorArray(dtype=dtypes.float32, size=1) 65 ta = tensor_array_ops.TensorArray(dtype=dtypes.float32, size=1) 85 ta = tensor_array_ops.TensorArray(dtype=dtypes.float32, size=1) 225 ta = tensor_array_ops.TensorArray(dtype=dtypes.float32, size=1) 247 ta = tensor_array_ops.TensorArray(dtype=dtypes.float32, size=1)
|
/aosp_15_r20/external/tensorflow/tensorflow/python/kernel_tests/data_structures/ |
H A D | tensor_array_ops_test.py | 62 return tensor_array_ops.TensorArray( 83 ta = tensor_array_ops.TensorArray( 104 ta = tensor_array_ops.TensorArray( 133 ta = tensor_array_ops.TensorArray( 150 ta = tensor_array_ops.TensorArray( 159 ta = tensor_array_ops.TensorArray( 175 tensor_array_ops.TensorArray( 189 ta = tensor_array_ops.TensorArray( 217 ta = tensor_array_ops.TensorArray( 233 ta = tensor_array_ops.TensorArray( [all …]
|
/aosp_15_r20/external/tensorflow/tensorflow/tools/consistency_integration_test/ |
H A D | tensor_array_tests.py | 65 ta = tf.TensorArray(size=3, dtype=tf.int32, element_shape=()) 105 ta = tf.TensorArray(tf.float32, num_rows) 113 ta1 = tf.TensorArray(tf.float32, num_rows) 145 ta = tf.TensorArray(dtype=tf.float32, dynamic_size=True, size=0) 167 ta = tf.TensorArray(dtype=tf.float32, dynamic_size=True, size=0) 194 ta = tf.TensorArray(dtype=tf.float32, dynamic_size=True, size=0) 202 a_ta = tf.TensorArray(dtype=tf.float32, dynamic_size=True, size=0) 224 ta0 = tf.TensorArray(dtype=tf.float32, dynamic_size=True, size=0) 228 ta1 = tf.TensorArray(dtype=tf.float32, dynamic_size=True, size=0) 281 ta = tf.TensorArray(x.dtype, tf.shape(x)[0])
|
/aosp_15_r20/external/tensorflow/tensorflow/python/data/util/ |
H A D | structure_test.py | 54 ("TensorArray", lambda: tensor_array_ops.TensorArray( 114 ("TensorArray", lambda: tensor_array_ops.TensorArray( 116 tensor_array_ops.TensorArray( 118 tensor_array_ops.TensorArray( 121 tensor_array_ops.TensorArray( 123 tensor_array_ops.TensorArray( 192 ("TensorArray", lambda: tensor_array_ops.TensorArray( 194 lambda: tensor_array_ops.TensorArray( 196 lambda: tensor_array_ops.TensorArray( 256 ("TensorArray", lambda: tensor_array_ops.TensorArray( [all …]
|
/aosp_15_r20/external/tensorflow/tensorflow/compiler/xla/g3doc/ |
H A D | known_issues.md | 20 ## TensorArray TF/XLA interconversion is not supported 25 XLA supports `tf.TensorArray`. However, the _interconversion_ between TF and XLA 27 `TensorArray` is used inside the compiled block, but the derivative is taken 35 number of elements. This could also happen if you're using a TensorArray in a 41 intermediate results in a `TensorArray`, but XLA only supports bounded 42 `TensorArray`s. 48 ## Dynamic `tf.TensorArray` is not supported 50 Writes into `tf.TensorArray(..., dynamic_size=True)` are not compilable with
|
/aosp_15_r20/external/tensorflow/tensorflow/core/kernels/ |
H A D | tensor_array_ops.cc | 76 Status GetTensorArray(OpKernelContext* ctx, TensorArray** tensor_array) { in GetTensorArray() 121 TensorArray* output_tensor_array; in Compute() 151 TensorArray** output_tensor_array) = 0; 184 TensorArray** output_tensor_array) override { in CreateTensorArray() 201 TensorArray::tensor_array_counter.fetch_add(1)); in CreateTensorArray() 207 TensorArray* tensor_array = new TensorArray( in CreateTensorArray() 281 TensorArray** output_tensor_array) override { in CreateTensorArray() 307 TensorArray* tensor_array; in CreateTensorArray() 354 tensor_array_output_handle](TensorArray** ret) -> Status { in CreateTensorArray() 355 *ret = new TensorArray( in CreateTensorArray() [all …]
|
H A D | tensor_array.h | 130 class TensorArray : public ResourceBase { 138 TensorArray(const string& key, const DataType& dtype, const Tensor& handle, in TensorArray() function 334 Status CopyShapesFrom(TensorArray* rhs, const TensorShape* shape_to_prepend); 347 return ctx->step_container()->MakeResourceHandle<TensorArray>( in resource_handle() 440 Status TensorArray::LockedWriteOrAggregate(OpKernelContext* ctx, in LockedWriteOrAggregate() 542 Status TensorArray::LockedRead(OpKernelContext* ctx, const int32_t index, in LockedRead()
|
H A D | tensor_array.cc | 82 std::atomic<int64_t> TensorArray::tensor_array_counter{0}; 84 Status TensorArray::CopyShapesFrom(TensorArray* rhs, in CopyShapesFrom()
|
/aosp_15_r20/external/tensorflow/tensorflow/python/ops/ |
H A D | tensor_array_grad.py | 104 g = (tensor_array_ops.TensorArray(dtype=dtype, handle=handle, flow=flow, 136 g = (tensor_array_ops.TensorArray(dtype=dtype, handle=handle, flow=flow, 167 g = (tensor_array_ops.TensorArray(dtype=dtype, handle=handle, flow=flow, 197 g = (tensor_array_ops.TensorArray(dtype=dtype, handle=handle, flow=flow, 228 g = (tensor_array_ops.TensorArray(dtype=dtype, handle=handle, flow=flow, 258 g = (tensor_array_ops.TensorArray(dtype=dtype, handle=handle, flow=flow,
|
H A D | tensor_array_ops_test.py | 32 values = tensor_array_ops.TensorArray( 50 values = tensor_array_ops.TensorArray( 63 values = tensor_array_ops.TensorArray( 76 values = tensor_array_ops.TensorArray(
|
H A D | tensor_array_ops.py | 245 g = TensorArray( 957 class TensorArray: class 1306 impl = (old_ta._implementation if isinstance(old_ta, TensorArray) else old_ta) 1319 new_ta = TensorArray( 1352 value_type = property(lambda self: TensorArray) 1427 if not isinstance(value, TensorArray): 1444 ret = TensorArray( 1454 if not isinstance(value, TensorArray): 1473 return TensorArray 1479 TensorArray, TensorArraySpec.from_value, allow_subclass=True)
|
/aosp_15_r20/external/tensorflow/tensorflow/python/autograph/operators/ |
H A D | data_structures.py | 92 l = tensor_array_ops.TensorArray( 185 if isinstance(list_, tensor_array_ops.TensorArray): 255 if isinstance(list_, tensor_array_ops.TensorArray): 319 if isinstance(list_, tensor_array_ops.TensorArray):
|
/aosp_15_r20/external/tensorflow/tensorflow/python/ops/parallel_for/ |
H A D | control_flow_ops_test.py | 1000 ta = tensor_array_ops.TensorArray( 1011 ta = tensor_array_ops.TensorArray( 1022 t = tensor_array_ops.TensorArray(dtypes.int32, 10, clear_after_read=False) 1031 out1 = tensor_array_ops.TensorArray( 1036 out2 = tensor_array_ops.TensorArray( 1046 ta1 = tensor_array_ops.TensorArray(dtypes.int32, 2).write(0, 1048 ta2 = tensor_array_ops.TensorArray(dtypes.int32, 1).write(0, 1) 1058 ta1 = tensor_array_ops.TensorArray(dtypes.int32, 1062 ta2 = tensor_array_ops.TensorArray(dtypes.int32, 1072 ta1 = tensor_array_ops.TensorArray( [all …]
|