/external/tensorflow/tensorflow/python/keras/layers/ |
D | kernelized_test.py | 79 rff_layer = kernel_layers.RandomFourierFeatures(output_dim=10, scale=3.0) 95 _ = kernel_layers.RandomFourierFeatures(output_dim=-3, scale=2.0) 107 _ = kernel_layers.RandomFourierFeatures(output_dim=10, scale=0.0) 111 rff_layer = kernel_layers.RandomFourierFeatures(output_dim=10, scale=3.0) 122 output_dim=10, 126 self.assertEqual(rff_layer.output_dim, 10) 136 output_dim=10, 152 kernel_layers.RandomFourierFeatures(output_dim=4, name='rff')(inputs) 153 kernel_layers.RandomFourierFeatures(output_dim=10, scale=2.0)(inputs) 158 output_dim=7, name='random_fourier_features', trainable=True) [all …]
|
D | embeddings.py | 93 output_dim, argument 106 if input_dim <= 0 or output_dim <= 0: 109 input_dim, output_dim)) 123 self.output_dim = output_dim 143 shape=(self.input_dim, self.output_dim), 151 shape=(self.input_dim, self.output_dim), 168 return input_shape + (self.output_dim,) 187 return (input_shape[0],) + tuple(in_lens) + (self.output_dim,) 206 'output_dim': self.output_dim,
|
D | kernelized.py | 152 output_dim, argument 158 if output_dim <= 0: 161 output_dim)) 172 self.output_dim = output_dim 193 self.kernel_initializer, shape=(input_dim, self.output_dim)) 197 shape=(input_dim, self.output_dim), 204 shape=(self.output_dim,), 236 return input_shape[:-1].concatenate(self.output_dim) 243 'output_dim': self.output_dim,
|
D | embeddings_test.py | 84 layer = keras.layers.Embedding(output_dim=2, input_dim=2) 94 keras.layers.Embedding(input_dim=0, output_dim=1) 97 keras.layers.Embedding(input_dim=1, output_dim=0) 101 l = keras.layers.Embedding(output_dim=2, input_dim=2) 115 output_dim=2, 138 layer = keras.layers.Embedding(input_dim=5, output_dim=2) 154 layer = keras.layers.Embedding(input_dim=5, output_dim=2)
|
D | cudnn_recurrent_test.py | 407 output_dim = 2 411 target_dim = 2 * output_dim if mode == 'concat' else output_dim 418 rnn(output_dim), merge_mode=mode, input_shape=(None, dim))) 431 rnn(output_dim, return_sequences=True), 434 model.add(keras.layers.Bidirectional(rnn(output_dim), merge_mode=mode)) 441 rnn(output_dim), merge_mode=mode)( 450 rnn(output_dim, stateful=True), merge_mode=mode)(
|
D | wrappers_test.py | 541 output_dim = 2 544 target_dim = 2 * output_dim if mode == 'concat' else output_dim 551 rnn(output_dim), merge_mode=mode, input_shape=(timesteps, dim))) 586 output_dim = 2 592 rnn(output_dim), input_shape=(timesteps, dim))) 605 output_dim = 2 610 target_dim = 2 * output_dim if mode == 'concat' else output_dim 616 rnn(output_dim, return_sequences=True), 619 model.add(keras.layers.Bidirectional(rnn(output_dim), merge_mode=mode)) 626 rnn(output_dim), merge_mode=mode)(inputs) [all …]
|
/external/tensorflow/tensorflow/python/keras/saving/ |
D | save_weights_test.py | 90 output_dim = 3 104 (keras.layers.Conv1D(output_dim, size, use_bias=False)), 105 [np.random.random((output_dim, input_dim, size, 1))], 109 (keras.layers.Conv2D(output_dim, size, 111 [np.random.random((output_dim, input_dim, size, size))], 115 (keras.layers.Conv2DTranspose(output_dim, size, 118 [np.random.random((output_dim, input_dim, size, size))], 122 (keras.layers.Conv2DTranspose(output_dim, size, 125 [np.random.random((size, size, input_dim, output_dim))], 129 (keras.layers.Conv3D(output_dim, size, [all …]
|
/external/tensorflow/tensorflow/compiler/xla/client/lib/ |
D | conv_grad_size_util.cc | 70 TF_ASSIGN_OR_RETURN(SpatialDimensionOutputSizeAndPadding output_dim, in ConvGradExtractAndVerifyDimension() 73 if (output_size != output_dim.output_size) { in ConvGradExtractAndVerifyDimension() 76 output_size, ", computed = ", output_dim.output_size, in ConvGradExtractAndVerifyDimension() 83 dim.output_size = (output_dim.output_size - 1) * stride + 1; in ConvGradExtractAndVerifyDimension() 85 dim.pad_before = effective_filter_size - 1 - output_dim.pad_before; in ConvGradExtractAndVerifyDimension()
|
/external/tensorflow/tensorflow/core/util/ |
D | bcast.h | 202 int output_dim = -1; in BCastList() local 206 output_dim = -1; in BCastList() 217 if (!output_dim_set || copy[i][j] == output_dim) { in BCastList() 218 output_dim = copy[i][j]; in BCastList() 226 output_.push_back(output_dim_set ? output_dim : 1); in BCastList() 262 result_.back() = mul_dims(result_.back(), output_dim); in BCastList() 266 mul_dims(bcast_[i].back(), current_is_one[i] ? output_dim : 1); in BCastList() 272 result_.push_back(output_dim); in BCastList() 275 bcast_[i].push_back(current_is_one[i] ? output_dim : 1); in BCastList()
|
/external/tensorflow/tensorflow/lite/kernels/internal/ |
D | batch_to_space_nd_test.cc | 25 int input_dim, int output_dim) { in GetIndexRange() argument 29 output_dim, &index_start, &index_end); in GetIndexRange()
|
/external/tensorflow/tensorflow/lite/delegates/xnnpack/ |
D | pad_tester.cc | 41 int32_t output_dim = InputShape()[i]; in OutputShape() local 43 output_dim += InputPrePaddings()[i]; in OutputShape() 46 output_dim += InputPostPaddings()[i]; in OutputShape() 48 output_shape.push_back(output_dim); in OutputShape()
|
D | reshape_tester.h | 46 for (int32_t output_dim : output_shape) { in OutputShape() local 47 EXPECT_GT(output_dim, 0); in OutputShape()
|
/external/tensorflow/tensorflow/core/ops/ |
D | image_ops.cc | 227 DimensionHandle output_dim; in CombinedNMSShapeFn() local 230 TF_RETURN_IF_ERROR(c->MakeDimForScalarInput(3, &output_dim)); in CombinedNMSShapeFn() 231 if (c->ValueKnown(output_dim) && c->Value(output_dim) <= 0) { in CombinedNMSShapeFn() 241 output_size = c->Value(output_dim); in CombinedNMSShapeFn() 248 output_size = std::min(c->Value(output_dim), in CombinedNMSShapeFn() 1026 DimensionHandle output_dim; in __anon30b5031d1b02() local 1027 TF_RETURN_IF_ERROR(c->MakeDimForScalarInput(2, &output_dim)); in __anon30b5031d1b02() 1028 c->set_output(0, c->MakeShape({output_dim})); in __anon30b5031d1b02() 1053 DimensionHandle output_dim; in __anon30b5031d1c02() local 1054 TF_RETURN_IF_ERROR(c->MakeDimForScalarInput(2, &output_dim)); in __anon30b5031d1c02() [all …]
|
/external/tensorflow/tensorflow/python/keras/benchmarks/keras_examples_benchmarks/ |
D | antirectifier_benchmark_test.py | 138 output_dim = input_shape[-1] 140 shape=(output_dim * 2, output_dim),
|
D | text_classification_transformer_benchmark_test.py | 224 input_dim=vocab_size, output_dim=embed_dim) 226 input_dim=maxlen, output_dim=embed_dim)
|
/external/tensorflow/tensorflow/compiler/xla/service/ |
D | dynamic_padder.cc | 355 for (int64 output_dim : output_dims) { in RewriteDynamicReshapeSplitInput() local 356 reshaped_dims.push_back(reshape->shape().dimensions(output_dim)); in RewriteDynamicReshapeSplitInput() 379 const int64 output_dim = output_dims[i]; in RewriteDynamicReshapeSplitInput() local 380 HloInstruction* dynamic_size = output_dynamic_dims[output_dim]; in RewriteDynamicReshapeSplitInput() 472 for (int64 output_dim : output_dims) { in RewriteDynamicReshapeSplitInput() local 474 dynamic_dimension_inference->GetDynamicSize(reshape, {}, output_dim); in RewriteDynamicReshapeSplitInput() 479 output_dim)); in RewriteDynamicReshapeSplitInput() 557 int64 output_dim, absl::Span<HloInstruction*> input_dynamic_dims, in RewriteDynamicReshapeCombineInput() argument 569 ShapeUtil::MakeShape(xla::S32, {output_shape.dimensions(output_dim)}); in RewriteDynamicReshapeCombineInput() 655 if (i != output_dim) { in RewriteDynamicReshapeCombineInput() [all …]
|
D | indexed_array_analysis.cc | 215 for (int64 output_dim : output_dims) { in FoldGatherOfGather() local 216 simulated_index.insert(simulated_index.begin() + output_dim, in FoldGatherOfGather() 537 for (int64 output_dim : operand->output_dims()) { in ReshapeToAddDegenerateDims() local 538 output_dims_bitvector[output_dim] = true; in ReshapeToAddDegenerateDims() 564 for (int64 output_dim : new_output_dims) { in ReshapeToAddDegenerateDims() local 565 EraseAt(&new_source_shape_dims, output_dim); in ReshapeToAddDegenerateDims() 711 int64 output_dim = scalar_indexed->output_dims()[i]; in FoldReshapeOfGatherNoDegenerateDims() local 713 reshape_passthrough_dims, output_dim); in FoldReshapeOfGatherNoDegenerateDims() 877 auto is_broadcasted_dim = [&](int64 output_dim) { in ComputeArrayForElementwiseBinaryOp() argument 878 return absl::c_find(broadcast_dims, output_dim) == broadcast_dims.end(); in ComputeArrayForElementwiseBinaryOp()
|
/external/tensorflow/tensorflow/python/debug/examples/v2/ |
D | debug_mnist_v2.py | 164 def get_dense_weights(input_dim, output_dim): argument 168 kernel = tf.Variable(initial_kernel([input_dim, output_dim])) 169 bias = tf.Variable(tf.constant(0.1, shape=[output_dim]))
|
/external/tensorflow/tensorflow/python/debug/examples/v1/ |
D | debug_mnist_v1.py | 162 def nn_layer(input_tensor, input_dim, output_dim, layer_name, act=tf.nn.relu): argument 168 weights = weight_variable([input_dim, output_dim]) 170 biases = bias_variable([output_dim])
|
/external/XNNPACK/test/ |
D | constant-pad-operator-tester.h | 84 inline size_t output_dim(size_t i) const { in output_dim() function 91 elements *= output_dim(i); in num_output_elements() 126 output_dims[XNN_MAX_TENSOR_DIMS - num_dims() + i] = output_dim(i); in TestX32()
|
/external/tensorflow/tensorflow/python/keras/distribute/ |
D | keras_embedding_model_correctness_test.py | 42 word_embed = keras.layers.Embedding(input_dim=20, output_dim=10)(word_ids) 100 output_dim=10,
|
D | keras_stateful_lstm_model_correctness_test.py | 64 word_embed = keras.layers.Embedding(input_dim=20, output_dim=10)(word_ids)
|
/external/tensorflow/tensorflow/python/keras/utils/ |
D | multi_gpu_utils_test.py | 59 output_dim = 1 71 model.add(keras.layers.Dense(output_dim)) 74 y = np.random.random((num_samples, output_dim))
|
/external/tensorflow/tensorflow/python/keras/ |
D | backend_test.py | 1184 output_dim = 3 1190 (num_samples, output_dim)).astype(np.float32) 1191 w_i_val = np.random.random((input_dim, output_dim)).astype(np.float32) 1192 w_o_val = np.random.random((output_dim, output_dim)).astype(np.float32) 1250 self.assertEqual(last_output.shape.as_list(), [num_samples, output_dim]) 1252 [num_samples, timesteps, output_dim]) 1254 self.assertEqual(state.shape.as_list(), [num_samples, output_dim]) 1294 output_dim = 3 1300 (num_samples, output_dim)).astype(np.float32) 1301 w_i_val = np.random.random((input_dim, output_dim)).astype(np.float32) [all …]
|
/external/tensorflow/tensorflow/python/kernel_tests/ |
D | broadcast_to_ops_test.py | 58 for output_dim in range(input_dim, 6): 61 output_shape = [2] * output_dim
|