| /external/tensorflow/tensorflow/contrib/tpu/python/tpu/ |
| tpu_test.py | 57 outputs = convolutional.conv2d(
|
| /external/tensorflow/tensorflow/python/kernel_tests/ |
| conv_ops_test.py | 206 conv = nn_ops.conv2d(
241 conv = nn_ops.conv2d(
289 computed = nn_ops.conv2d(
[all...] |
| /external/tensorflow/tensorflow/contrib/fused_conv/python/ops/ |
| fused_conv2d_bias_activation_benchmark.py | 15 """Benchmark for fused conv2d bias and activation op."""
34 """builds a graph containing a sequence of conv2d operations.
44 num_iters: number of iterations to run conv2d.
62 conv2d_out = nn_ops.conv2d(
69 conv2d_out = nn_ops.conv2d(
79 """builds a graph containing a sequence of conv2d operations.
89 num_iters: number of iterations to run conv2d.
132 """Benchmark conv2d!"""
147 num_iters: number of iterations to run conv2d.
227 print("fused conv2d bias activation benchmark using resnet50's shapes:"
[all...] |
| fused_conv2d_bias_activation_op_test.py | 15 """Functional tests for fused conv2d bias and activation operation."""
216 ref_conv_output = nn_ops.conv2d(
642 conv_result = nn_ops.conv2d(
[all...] |
| /external/tensorflow/tensorflow/examples/learn/ |
| text_classification_character_cnn.py | 54 conv1 = tf.layers.conv2d(
71 conv2 = tf.layers.conv2d(
|
| /external/tensorflow/tensorflow/contrib/quantize/python/ |
| fold_batch_norms_test.py | 39 conv2d = layers.conv2d variable
70 """Tests folding cases: inputs -> Conv2d with batch norm -> Relu*.
90 node = conv2d(
117 self.assertEqual(folded_conv.type, 'Conv2D')
137 """Tests folding cases: inputs -> Conv2d with batch norm -> Relu*.
159 node = conv2d(
186 self.assertEqual(folded_conv.type, 'Conv2D')
368 node = conv2d(
|
| quantize_parameterized_test.py | 34 conv2d = layers.conv2d variable
63 """Tests quantization: inputs -> Conv2d no batch norm -> Activation.
81 node = conv2d(inputs, out_depth, [5, 5], stride=stride, padding='SAME',
104 output_op_name = scope + '/Conv2D'
327 """Tests quantization: inputs -> Conv2d with batch norm -> Activation.
345 node = conv2d(
|
| quantize_graph_test.py | 219 conv = layers.conv2d(
|
| fold_batch_norms.py | 40 Folding only affects the following layers: Conv2D, fully connected, depthwise
63 Folding only affects the following layers: Conv2D, fully connected, depthwise
160 'Conv2D|DepthwiseConv2dNative|MatMul',
380 if layer_op.type == 'Conv2D':
381 return nn_ops.conv2d(
420 Folding only affects the following layers: Conv2D, fully connected, depthwise
546 mul is cloned into mul_fold, Conv2D or MatMul, or DepthwiseConv2d is cloned
604 elif op_below.type in ['Conv2D', 'MatMul']:
662 'Conv2D': self._CloneConv2d,
679 return nn_ops.conv2d(
[all...] |
| /external/tensorflow/tensorflow/contrib/slim/python/slim/nets/ |
| resnet_v1_test.py | 97 y1 = layers.conv2d(x, 1, [3, 3], stride=1, scope='Conv')
109 y4 = layers.conv2d(x, 1, [3, 3], stride=2, scope='Conv')
134 y1 = layers.conv2d(x, 1, [3, 3], stride=1, scope='Conv')
148 y4 = layers.conv2d(x, 1, [3, 3], stride=2, scope='Conv')
161 with arg_scope([layers.conv2d], outputs_collections='end_points'):
|
| resnet_v2_test.py | 97 y1 = layers.conv2d(x, 1, [3, 3], stride=1, scope='Conv')
109 y4 = layers.conv2d(x, 1, [3, 3], stride=2, scope='Conv')
134 y1 = layers.conv2d(x, 1, [3, 3], stride=1, scope='Conv')
151 y4 = layers.conv2d(x, 1, [3, 3], stride=2, scope='Conv')
164 with arg_scope([layers.conv2d], outputs_collections='end_points'):
|
| /external/tensorflow/tensorflow/python/tools/ |
| optimize_for_inference_test.py | 140 conv_op = nn_ops.conv2d(
186 conv_op = nn_ops.conv2d(
239 nn_ops.conv2d(
254 self.assertNotEqual("Conv2D", node.op)
268 nn_ops.conv2d(
283 self.assertNotEqual("Conv2D", node.op)
297 nn_ops.conv2d(
312 self.assertNotEqual("Conv2D", node.op)
|
| /external/tensorflow/tensorflow/examples/speech_commands/ |
| models.py | 176 [Conv2D]<-(weights)
184 [Conv2D]<-(weights)
227 first_conv = tf.nn.conv2d(fingerprint_4d, first_weights, [1, 1, 1, 1],
246 second_conv = tf.nn.conv2d(max_pool, second_weights, [1, 1, 1, 1],
285 [Conv2D]<-(weights)
335 first_conv = tf.nn.conv2d(fingerprint_4d, first_weights, [
|
| /external/tensorflow/tensorflow/python/layers/ |
| convolutional_test.py | 44 conv_layers.conv2d(images, 32, 3, data_format='invalid')
50 conv_layers.conv2d(images, 32, 3, strides=(1, 2, 3))
53 conv_layers.conv2d(images, 32, 3, strides=None)
59 conv_layers.conv2d(images, 32, (1, 2, 3))
62 conv_layers.conv2d(images, 32, None)
67 layer = conv_layers.Conv2D(32, [3, 3], activation=nn_ops.relu)
69 self.assertEqual(output.op.name, 'conv2d/Relu')
78 output = conv_layers.conv2d(images, 32, [3, 3], activation=nn_ops.relu)
85 layer = conv_layers.Conv2D(32, 3)
95 layer = conv_layers.Conv2D(32, [3, 3], data_format='channels_first'
[all...] |
| layers.py | 23 @@Conv2D
45 @@conv2d
92 from tensorflow.python.layers.convolutional import Conv2D
102 from tensorflow.python.layers.convolutional import conv2d
|
| /external/tensorflow/tensorflow/contrib/tensorrt/test/ |
| test_tftrt.py | 23 # tf.placeholder, tf.constant, tf.nn.conv2d etc but
49 conv = nn.conv2d(
|
| /external/tensorflow/tensorflow/contrib/model_pruning/examples/cifar10/ |
| cifar10_pruning.py | 192 conv = tf.nn.conv2d(
214 conv = tf.nn.conv2d(
|
| /external/tensorflow/tensorflow/python/grappler/ |
| memory_optimizer_test.py | 116 after_conv = nn.conv2d(current_activation, conv_filter, [1, 1, 1, 1],
228 after_conv = nn.conv2d(current_activation, conv_filter, [1, 1, 1, 1],
|
| cost_analyzer_test.py | 70 conv = nn_ops.conv2d(image, w, strides=[1, 1, 1, 1], padding="SAME")
91 self.assertTrue(b"Conv2D" in report)
95 for op_type in [b"MatMul", b"Conv2D", b"Conv2DBackpropFilter"]:
|
| /external/tensorflow/tensorflow/contrib/metrics/python/ops/ |
| histogram_ops.py | 242 result = nn_ops.conv2d(x, h, [1, 1, 1, 1], 'SAME')
|
| /external/tensorflow/tensorflow/python/keras/_impl/keras/ |
| backend_test.py | 766 y = keras.backend.conv2d(x, k,
772 y = keras.backend.conv2d(x, k, strides=(1, 1),
778 y = keras.backend.conv2d(x, k, strides=(1, 1),
784 y = keras.backend.conv2d(x, k, strides=(2, 2),
788 y = keras.backend.conv2d(x, k, (2, 2),
791 y = keras.backend.conv2d(x, k, (2, 2),
794 y = keras.backend.conv2d(x, k, (2, 2, 2))
[all...] |
| /external/tensorflow/tensorflow/core/graph/ |
| mkl_layout_pass.cc | 55 // gradient ops of Conv2D+AddBias. Gradient op of both the Conv2D and
57 // Conv2D-specific BiasAddGrad, and MatMul-specific BiasAddGrad.
63 // Currently, we merge Conv2D+AddBias together. Consider Conv2D and BiasAdd as:
65 // O = Conv2D(A, B)
74 // - The merge for Conv2D and BiasAdd happens when the output of Conv2D _only_
120 // and Mkl tensors. E.g., assume an op 'Conv2D' that takes (A, B) as
255 // we consider it Conv2D context; if it is MatMul, then it is MatMul context
465 string conv2d; member in struct:tensorflow::MklLayoutRewritePass::__anon39629
[all...] |
| /external/tensorflow/tensorflow/python/keras/_impl/keras/layers/ |
| convolutional_recurrent.py | 522 conv_out = K.conv2d(
534 conv_out = K.conv2d(
|
| /external/tensorflow/tensorflow/contrib/gan/python/eval/python/ |
| sliced_wasserstein_impl.py | 61 conv_out = nn_ops.conv2d(xt, gaussian_filter * gain, [1] * 4, 'VALID')
|
| /external/tensorflow/tensorflow/compiler/tests/ |
| depthwise_conv_op_test.py | 40 # Use a custom implementation of depthwise conv2d using slicing.
51 convs.append(nn_ops.conv2d(input_slice, filter_slice,
|
