experiments/STT-tensorflow
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
STT-tensorflow/tensorflow/core/ops
History
TensorFlower Gardener e264e71a44 Merge pull request #40480 from yongtang:40471-equal-output-shapes-autograph 
PiperOrigin-RevId: 317396929
Change-Id: Id5ce5a139055d367e97ff39fc65cbe68af3146b6
2020-06-19 16:13:47 -07:00
..
compat
Update ops-related pbtxt files.
2020-06-18 21:22:33 -07:00
array_grad_test.cc
Add broadcast optimization rewrite to Grappler. This generalizes the rewrites for algebraically neutral element simplications (e.g. x + zeros(shape) => x ) to handle the cases where the output shape differs from the input shape of the non-trivial argument. This pattern is sometimes used in tensorflow for broadcasting.
2019-04-12 13:20:08 -07:00
array_grad.cc
Add broadcast optimization rewrite to Grappler. This generalizes the rewrites for algebraically neutral element simplications (e.g. x + zeros(shape) => x ) to handle the cases where the output shape differs from the input shape of the non-trivial argument. This pattern is sometimes used in tensorflow for broadcasting.
2019-04-12 13:20:08 -07:00
array_ops_test.cc
Allow targets to pass with -fsanitize=null under --config=asan.
2020-04-16 12:14:00 -07:00
array_ops.cc
Add Registration for non-stateful scatter_nd_min and scatter_nd_max.
2020-06-03 11:05:54 -07:00
audio_ops.cc
…
batch_ops.cc
Introduce op attribute 'enable_large_batch_splitting' and keep default behavior consistent.
2020-05-27 15:18:30 -07:00
bitwise_ops.cc
Remove .SetIsCommutative() from ops that are not actually commutative w.r.t. all their inputs.
2019-04-19 14:43:40 -07:00
boosted_trees_ops.cc
Fix failing ASan test due to out of boundary memory access
2020-02-01 05:48:32 -08:00
candidate_sampling_ops_test.cc
…
candidate_sampling_ops.cc
…
checkpoint_ops.cc
…
clustering_ops.cc
Move clustering ops to core.
2019-01-10 17:39:50 -08:00
collective_ops.cc
Add timeout to collective ops to detect deadlocks.
2020-05-29 15:37:42 -07:00
control_flow_ops_test.cc
Adds a lowering from Case to _SwitchN+Merge.
2019-05-30 18:28:08 -07:00
control_flow_ops.cc
Adds a lowering from Case to _SwitchN+Merge.
2019-05-30 18:28:08 -07:00
count_ops.cc
Automated g4 rollback of changelist 311477582.
2020-05-14 08:42:30 -07:00
ctc_ops_test.cc
…
ctc_ops.cc
Add changes to support cuDNN CTC loss
2019-09-06 15:49:54 -07:00
cudnn_rnn_ops_test.cc
Merge pull request #27756 from houtoms:lstmp_upstream
2019-06-10 22:47:24 -07:00
cudnn_rnn_ops.cc
add cudnn lstm projection (lstmp)
2019-04-10 21:43:35 -07:00
data_flow_ops_test.cc
…
data_flow_ops.cc
Fix static shape computation of dynamic_stitch when all tensors are empty.
2019-10-11 14:12:41 -07:00
dataset_ops.cc
[tf.data] Move cardinality op out of experimental
2020-06-03 12:12:24 -07:00
debug_ops.cc
[tfdbg2] Add tfdbg_run_id to metadata of data dumps
2020-06-15 13:52:52 -07:00
decode_proto_ops.cc
…
encode_proto_ops.cc
…
experimental_dataset_ops.cc
[tf.data] Implementation of tf.data.experimental.save and tf.data.experimental.load. The former makes it possible to save the output of a dataset, while the latter makes it possible to load a previously saved data.
2020-06-11 15:37:49 -07:00
function_ops.cc
Enable setting shape of _Arg tensor for shape of resource args
2020-04-28 06:40:58 -07:00
functional_grad.cc
…
functional_ops_test.cc
Enable setting shape of _Arg tensor for shape of resource args
2020-04-28 06:40:58 -07:00
functional_ops.cc
Add a function to dynamic choose and execute the proper implementation based on underlying device placement
2020-05-28 09:54:31 -07:00
image_ops_test.cc
Update the api_def
2019-01-28 21:58:03 +00:00
image_ops.cc
Use ExtractGlimpseV2 and ExtractGlimpse to make sure C++ kernel is backward compatible
2020-05-12 09:31:29 -07:00
io_ops_test.cc
…
io_ops.cc
Updated the majority of string tensor accessors to use tstring type.
2019-08-07 11:46:35 -07:00
linalg_ops_test.cc
Add Broadcasted Matrix Triangular Solve.
2020-01-27 20:43:43 -08:00
linalg_ops.cc
Add banded triangular solve op.
2020-06-18 10:03:35 -07:00
list_ops.cc
Check handle_data vector size before inspecting first element in list_ops shape inference functions. The current behavior can segfault.
2020-06-02 09:50:09 -07:00
logging_ops.cc
PrintV2 shape inference should succeed if rank is unknown
2019-10-25 14:10:48 -07:00
lookup_ops.cc
Allowing tf.lookup tables to be initialized from a Dataset
2020-03-02 15:41:42 -08:00
manip_ops.cc
…
math_grad_test.cc
Add tf.math.xlog1py, a safe way to compute x * log1p(y)
2020-01-09 14:28:37 -08:00
math_grad.cc
Use explicit primitive types with strings::Printf
2020-03-08 20:58:24 -07:00
math_ops_test.cc
Move tf.equal shape inference test to math_ops_test.cc
2020-06-17 18:38:24 +00:00
math_ops.cc
Expand dtype support for Neg
2020-06-18 20:31:51 -07:00
mkl_array_ops.cc
[Intel MKL] Fix dequantize accuracy issue and re-enable this OP
2020-02-14 13:47:29 +08:00
mkl_nn_ops.cc
Merge branch 'master' into yang/bn_relu_fwd
2020-05-05 14:59:31 +08:00
nccl_ops.cc
…
nn_grad.cc
Add a bfloat16 sum reducer that uses float32 accumulators. Fix existing tests.
2020-06-01 10:21:39 -07:00
nn_ops_test.cc
Take #2: Improve Conv2DBackpropInput to take input_sizes as a 2D shape.
2020-03-26 16:33:53 -07:00
nn_ops.cc
Update operation summary for TF _FusedConv2D.
2020-06-10 12:04:48 -07:00
no_op.cc
…
ops.pbtxt
Update ops-related pbtxt files.
2020-06-18 21:22:33 -07:00
parsing_ops_test.cc
Extend tf.io.parse_sequence_example and tf.io.parse_single_sequence_example to handle ragged tensor features.
2019-10-03 05:29:58 -07:00
parsing_ops.cc
minor spelling tweaks
2020-02-27 15:42:16 +09:00
ragged_array_ops.cc
Add tf.ragged.cross and tf.ragged.cross_hashed operations. These ops generate feature crosses from a list of input tensors (similar to tf.sparse.cross), and return the result as a RaggedTensor. Inputs may be tf.RaggedTensors, tf.Tensors, or tf.SparseTensors.
2020-03-02 18:13:45 -08:00
ragged_conversion_ops.cc
Improve handling of variant-encoded ragged tensors with ragged_rank=0.
2019-10-02 13:39:32 -07:00
ragged_math_ops.cc
Update RaggedTensors to support int32 row_splits.
2019-04-24 19:39:31 -07:00
ragged_to_dense_util_test.cc
A ragged to dense op for directly calculating tensors.
2019-08-09 16:49:53 -07:00
ragged_to_dense_util.cc
Added benchmarks for RaggedTensor.to_dense. Also added a few new test cases and improved readability of a few error messages.
2019-09-20 11:14:42 -07:00
ragged_to_dense_util.h
A ragged to dense op for directly calculating tensors.
2019-08-09 16:49:53 -07:00
random_grad.cc
…
random_ops_test.cc
…
random_ops.cc
Fixes doc and error messages about non-scalar minval/maxval in integer uniform RNGs.
2020-03-05 17:28:06 -08:00
remote_fused_graph_ops_test.cc
…
remote_fused_graph_ops.cc
…
resource_variable_ops.cc
Roll forward https://github.com/tensorflow/tensorflow/commit/ed0e46c219efa62559901...
2020-03-09 16:40:34 -07:00
rnn_ops_test.cc
Migrated LSTMBlock* and GRUCell* ops from contrib/rnn to core
2019-06-06 03:09:43 -07:00
rnn_ops.cc
Added BlockLSTMV2 and BlockLSTMGradV2 with IFCO layout
2019-08-14 21:34:19 -07:00
rpc_ops.cc
…
scoped_allocator_ops.cc
minor spelling tweaks
2020-02-27 15:42:16 +09:00
script_ops.cc
Set py_func private executor in c++ instead of python.
2019-08-12 16:04:20 -07:00
sdca_ops.cc
…
sendrecv_ops.cc
Exported Send/Recv in tf.raw_ops
2019-09-23 23:49:49 -07:00
set_ops_test.cc
…
set_ops.cc
…
shape_function_test.cc
…
sparse_csr_matrix_ops_test.cc
Add CSRSparseMatrix ops.
2019-09-03 16:49:47 -07:00
sparse_csr_matrix_ops.cc
Add CSRSparseMatrix ops.
2019-09-03 16:49:47 -07:00
sparse_ops_test.cc
…
sparse_ops.cc
Add SparseCrossV2 which supports strong_hash with salt, and fingerprint doens't
2020-05-18 17:28:14 -07:00
special_math_ops.cc
Add Bessel functions to the public api:
2020-06-17 21:06:09 -07:00
spectral_ops_test.cc
…
spectral_ops.cc
Add complex128 support to RFFT, RFFT2D, RFFT3D, IRFFT, IRFFT2D, and IRFFT3D.
2019-09-12 15:07:02 -07:00
state_ops_test.cc
Check that rank(updates) = rank(indices + params[1:]) in resource_scatter_update, to match behavior of V1 scatter_update and shape function.
2020-06-11 20:32:05 -07:00
state_ops.cc
Add Registration for non-stateful scatter_nd_min and scatter_nd_max.
2020-06-03 11:05:54 -07:00
stateful_random_ops.cc
Fixes doc and error messages about non-scalar minval/maxval in integer uniform RNGs.
2020-03-05 17:28:06 -08:00
stateless_random_grad.cc
…
stateless_random_ops.cc
Add StatelessParameterizedTruncatedNormal sampler.
2020-06-18 10:28:38 -07:00
string_ops_test.cc
…
string_ops.cc
For Substr check pos and len rank equality only when their rank is known.
2019-10-17 17:10:41 -07:00
summary_ops.cc
Add tf.summary.experimental.write_raw_pb() to replace import_event()
2019-04-11 17:59:49 -07:00
tensor_forest_ops.cc
add tensor forest inference kernels
2018-11-13 12:25:05 -05:00
tpu_configuration_ops.cc
Don't shut down TPU chips if compilation fails in TF2.
2019-10-24 22:21:26 -07:00
tpu_cross_replica_ops.cc
Fix tpu_ops.all_to_all op output shape.
2019-08-09 19:30:25 -07:00
tpu_embedding_load_retrieve_ops.cc
Changed TPU embedding load and retrieve ops to checked-in generated code.
2020-03-19 13:52:03 -07:00
tpu_embedding_ops.cc
Add EnqueueTPUEmbeddingRaggedTensorBatch for RaggedTensor support.
2020-04-01 13:28:05 -07:00
tpu_functional_ops.cc
Adds config proto fields for enabling internal autotuning features.
2019-10-09 15:56:52 -07:00
tpu_heartbeat_ops.cc
Move TPU ops to TF core.
2019-02-13 17:35:30 -08:00
tpu_host_compute_ops.cc
Generate MLIR ops for TPU Host/Device communication for outside compilation.
2020-05-14 12:18:06 -07:00
tpu_infeed_ops.cc
Add ops to transform tensors to desired TPU device format as part of input pipeline.
2019-03-08 11:07:20 -08:00
tpu_ordinal_selector_ops.cc
Move TPU ops to TF core.
2019-02-13 17:35:30 -08:00
tpu_outfeed_ops.cc
Move TPU ops to TF core.
2019-02-13 17:35:30 -08:00
tpu_replication_ops.cc
Add an attribute "is_packed" to TPUReplicatedInput op which indicates whether the per-replica inputs are packed into one input.
2020-05-26 10:21:51 -07:00
training_ops_test.cc
minor spelling tweaks
2020-02-27 15:42:16 +09:00
training_ops.cc
Added CPU and GPU implementations of a modified version of FTRL, triggered using an attribute, that multiplies the "linear" accumulator by the learning rate. This enables a learning rate of zero to work correctly.
2020-04-07 12:11:41 -07:00
word2vec_ops.cc
…