| /external/tensorflow/tensorflow/contrib/memory_stats/python/kernel_tests/ |
| memory_stats_ops_test.py | 26 from tensorflow.python.ops import math_ops
64 c = math_ops.matmul(a, b)
65 d = math_ops.matmul(c, b)
80 c = math_ops.matmul(a, b)
|
| /external/tensorflow/tensorflow/python/ops/distributions/ |
| kullback_leibler.py | 24 from tensorflow.python.ops import math_ops
107 math_ops.logical_not(
108 math_ops.reduce_any(math_ops.is_nan(kl_t))),
|
| bernoulli.py | 25 from tensorflow.python.ops import math_ops
120 sample = math_ops.less(uniform, self.probs)
121 return math_ops.cast(sample, self.dtype)
130 event = math_ops.cast(event, self.logits.dtype)
146 return (-self.logits * (math_ops.sigmoid(self.logits) - 1) +
157 return math_ops.cast(self.probs > 0.5, self.dtype)
177 return (math_ops.sigmoid(a.logits) * delta_probs0
178 + math_ops.sigmoid(-a.logits) * delta_probs1)
|
| /external/tensorflow/tensorflow/python/training/ |
| proximal_gradient_descent.py | 23 from tensorflow.python.ops import math_ops
92 math_ops.cast(self._learning_rate_tensor, grad.dtype),
93 math_ops.cast(self._l1_regularization_strength_tensor, grad.dtype),
94 math_ops.cast(self._l2_regularization_strength_tensor, grad.dtype),
|
| /external/tensorflow/tensorflow/contrib/distributions/python/ops/ |
| cauchy.py | 29 from tensorflow.python.ops import math_ops
178 return math_ops.atan(self._z(x)) / np.pi + 0.5
181 return math_ops.log1p(2 / np.pi * math_ops.atan(self._z(x))) - np.log(2)
184 return -math_ops.log1p(math_ops.square(self._z(x)))
187 return np.log(np.pi) + math_ops.log(self.scale)
190 h = np.log(4 * np.pi) + math_ops.log(self.scale)
194 return self.loc + self.scale * math_ops.tan(np.pi * (p - 0.5))
|
| mvn_linear_operator.py | 25 from tensorflow.python.ops import math_ops
237 return array_ops.matrix_diag(math_ops.square(self.scale.diag_part()))
243 return math_ops.square(self.scale.diag_part())
254 return math_ops.abs(self.scale.diag_part())
257 return math_ops.sqrt(array_ops.matrix_diag_part(
260 return math_ops.sqrt(array_ops.matrix_diag_part(
302 # return math_ops.square(linalg_ops.norm(x, ord="fro", axis=[-2, -1]))
303 return math_ops.reduce_sum(math_ops.square(x), axis=[-2, -1])
337 - math_ops.cast(a.scale.domain_dimension_tensor(), a.dtype
[all...] |
| poisson.py | 27 from tensorflow.python.ops import math_ops
102 self._log_rate = math_ops.log(rate, name="log_rate")
108 self._rate = math_ops.exp(log_rate, name="rate")
147 return math_ops.log(self.cdf(x))
156 x = math_ops.floor(x)
157 return math_ops.igammac(1. + x, self.rate)
167 x = math_ops.floor(x)
168 return x * self.log_rate - math_ops.lgamma(1. + x)
180 return math_ops.floor(self.rate)
|
| poisson_lognormal.py | 30 from tensorflow.python.ops import math_ops
127 edges = math_ops.linspace(zero, 1., quadrature_size + 3)[1:-1]
134 math_ops.range(1, 1 + batch_ndims), [0]], axis=0)
149 value=1. / math_ops.cast(quadrature_size, dist.dtype))
280 logits=math_ops.log(self._quadrature_probs),
339 batch_size = math_ops.reduce_prod(self.batch_shape_tensor())
363 offset = math_ops.range(start=0,
376 return math_ops.reduce_logsumexp(
382 return math_ops.exp(
383 math_ops.reduce_logsumexp
[all...] |
| logistic.py | 30 from tensorflow.python.ops import math_ops
182 sampled = math_ops.log(uniform) - math_ops.log1p(-1. * uniform)
189 return math_ops.exp(self._log_prob(x))
195 return math_ops.sigmoid(self._z(x))
201 return math_ops.sigmoid(-self._z(x))
208 return math_ops.log(self.scale)
213 return 2 + math_ops.log(scale)
|
| onehot_categorical.py | 26 from tensorflow.python.ops import math_ops
197 logits_shape = array_ops.shape(math_ops.reduce_sum(logits, -1))
207 return math_ops.exp(self._log_prob(x))
210 return -math_ops.reduce_sum(
214 ret = math_ops.argmax(self.logits, axis=self._batch_rank)
221 ret = -math_ops.matmul(p[..., None], p[..., None, :])
234 math_ops.reduce_logsumexp(x, axis=[-1])),
254 return math_ops.reduce_sum(
|
| /external/tensorflow/tensorflow/python/grappler/ |
| cost_analyzer_test.py | 30 from tensorflow.python.ops import math_ops
45 c = math_ops.add_n([a, b], name="c")
46 d = math_ops.add_n([b, c], name="d")
77 y_conv = nn_ops.softmax(math_ops.matmul(h_conv_flat, w_fc) + b_fc)
79 cross_entropy = math_ops.reduce_mean(-math_ops.reduce_sum(
80 label * math_ops.log(y_conv), reduction_indices=[1]))
117 c = math_ops.add_n([a, b], name="c")
118 d = math_ops.add_n([b, c], name="d")
|
| /external/tensorflow/tensorflow/contrib/bayesflow/python/ops/ |
| hmc_impl.py | 35 from tensorflow.python.ops import math_ops
315 1 + array_ops.where(math_ops.equal(iter_, 0),
323 elems=math_ops.range(num_results), # iter_: used to choose burnin.
485 beta = (math_ops.cast(iter_ + 1, dtype)
486 / math_ops.cast(num_steps, dtype))
521 / math_ops.cast(num_steps, ais_weights.dtype))
819 random_positive = -math_ops.log(random_uniform)
845 acceptance_probs=math_ops.exp(math_ops.minimum(-energy_change, 0.)),
[all...] |
| custom_grad_impl.py | 27 from tensorflow.python.ops import math_ops
99 sum_x = math_ops.reduce_sum(x, axis=axis, name="sum_x")
|
| /external/tensorflow/tensorflow/contrib/gan/python/features/python/ |
| virtual_batchnorm_impl.py | 32 from tensorflow.python.ops import math_ops
64 y = math_ops.cast(x, dtypes.float32) if x.dtype == dtypes.float16 else x
67 shift = array_ops.stop_gradient(math_ops.reduce_mean(y, axes, keep_dims=True))
69 shifted_mean = math_ops.reduce_mean(y - shift, axes, keep_dims=True)
71 mean_squared = math_ops.reduce_mean(math_ops.square(y), axes, keep_dims=True)
76 return (math_ops.cast(mean, dtypes.float16),
77 math_ops.cast(mean_squared, dtypes.float16))
221 math_ops.square(self._ref_mean))
227 self._example_weight = 1. / (math_ops.to_float(ref_batch_size) + 1.
[all...] |
| /external/tensorflow/tensorflow/contrib/kfac/python/ops/ |
| optimizer.py | 30 from tensorflow.python.ops import math_ops
241 math_ops.reduce_sum(grad * pgrad)
244 return math_ops.reduce_sum(terms)
274 return math_ops.minimum(1.,
275 math_ops.sqrt(self._norm_constraint / sq_norm_up))
345 batch_size = math_ops.cast(
380 qmodel_change = 0.5 * math_ops.reduce_sum(sol * c)
405 math_ops.equal(m_22, 0.0), zero_prevupd_case, non_zero_prevupd_case)
487 return math_ops.add_n(
488 [math_ops.reduce_sum(elt1 * elt2) for elt1, elt2 in zip(list1, list2)]
[all...] |
| /external/tensorflow/tensorflow/contrib/nn/python/ops/ |
| fwd_gradients_test.py | 24 from tensorflow.python.ops import math_ops
32 y = math_ops.square(x)
|
| /external/tensorflow/tensorflow/contrib/quantization/ |
| __init__.py | 24 from tensorflow.contrib.quantization.python.math_ops import *
|
| /external/tensorflow/tensorflow/python/eager/ |
| graph_only_ops_test.py | 26 from tensorflow.python.ops import math_ops
40 y_tf = math_ops.square(x_tf)
|
| /external/tensorflow/tensorflow/python/kernel_tests/ |
| cross_grad_test.py | 23 from tensorflow.python.ops import math_ops
36 s = math_ops.cross(u, v)
|
| trace_op_test.py | 22 from tensorflow.python.ops import math_ops
34 tf_ans = math_ops.trace(x).eval()
|
| tensordot_op_test.py | 15 """Tests for tensorflow.ops.math_ops.matmul."""
27 from tensorflow.python.ops import math_ops
49 math_ops.tensordot(a, b, (a_axes, b_axes))
57 output = math_ops.tensordot(a_ph, b_ph, axes_ph)
71 math_ops.tensordot(a, b, axes_value)
74 math_ops.tensordot(a, b, [[0], [7]])
80 output = math_ops.tensordot(a_ph, b_ph, axes_ph)
102 tf_ans = math_ops.tensordot(tf_a, tf_b, axes_value).eval()
111 output = math_ops.tensordot(a, b, axes)
115 output = math_ops.tensordot(a, b, axes
[all...] |
| /external/tensorflow/tensorflow/python/ops/linalg/ |
| linear_operator_diag.py | 25 from tensorflow.python.ops import math_ops
215 math_ops.real(self._diag),
226 diag_term = math_ops.conj(self._diag) if adjoint else self._diag
232 return math_ops.reduce_prod(self._diag, reduction_indices=[-1])
235 return math_ops.reduce_sum(
236 math_ops.log(math_ops.abs(self._diag)), reduction_indices=[-1])
239 diag_term = math_ops.conj(self._diag) if adjoint else self._diag
|
| /external/tensorflow/tensorflow/python/ops/ |
| nn.py | 116 from tensorflow.python.ops.math_ops import sigmoid
117 from tensorflow.python.ops.math_ops import tanh
|
| nn_grad.py | 28 from tensorflow.python.ops import math_ops
242 math_ops.reduce_sum(grad_softmax * softmax, [1]), [-1, 1])) * softmax)
260 softmax = math_ops.exp(op.outputs[0])
261 return grad - math_ops.reduce_sum(grad, 1, keepdims=True) * softmax
348 reduction_dim_tensor = math_ops.range(array_ops.rank(received_grad) - 1)
349 return (received_grad, math_ops.reduce_sum(received_grad,
415 d2x = grad * dy / (math_ops.exp(-x) + 2.0 + math_ops.exp(x))
472 math_ops.matmul(grad_grad[:, None, :], softmax[:, :, None]), axis=1)) *
862 x = math_ops.cast(x, dtypes.float32
[all...] |
| /external/tensorflow/tensorflow/python/tools/ |
| strip_unused_test.py | 31 from tensorflow.python.ops import math_ops
46 wanted_input_node = math_ops.subtract(constant_node,
49 output_node = math_ops.multiply(
51 math_ops.add(output_node, 2.0, name="later_node")
110 input_node1 = math_ops.subtract(constant_node1, 3.0, name="input_node1")
111 input_node2 = math_ops.subtract(constant_node2, 5.0, name="input_node2")
112 output_node = math_ops.multiply(
114 math_ops.add(output_node, 2.0, name="later_node")
|
