- 10 8月, 2019 40 次提交
-
-
由 Anna R 提交于
PiperOrigin-RevId: 262675086
-
由 Xiao Yu 提交于
PiperOrigin-RevId: 262671025
-
由 A. Unique TensorFlower 提交于
PiperOrigin-RevId: 262668255
-
由 Ashwin Murthy 提交于
By default, all functions are exported including main. This flag will be used to support new use cases like conversion of speech models to tflite where we want to run specific inference graphs through the converter and export specific functions (e.g. Joint, Encoder etc.) to a separate flatbuffer. PiperOrigin-RevId: 262667934
-
由 A. Unique TensorFlower 提交于
PiperOrigin-RevId: 262667826
-
由 Revan Sopher 提交于
PiperOrigin-RevId: 262666528
-
由 River Riddle 提交于
This will allow for reusing the same pattern list, which may be costly to continually reconstruct, on multiple invocations. PiperOrigin-RevId: 262664599
-
由 Eileen Mao 提交于
PiperOrigin-RevId: 262664464
-
由 A. Unique TensorFlower 提交于
PiperOrigin-RevId: 262663630
-
由 A. Unique TensorFlower 提交于
PiperOrigin-RevId: 262660627
-
由 Brian Zhao 提交于
PiperOrigin-RevId: 262660551
-
由 Saurabh Saxena 提交于
PiperOrigin-RevId: 262655851
-
由 A. Unique TensorFlower 提交于
Eventually should be integrated into RaggedTensor.to_tensor. PiperOrigin-RevId: 262651646
-
由 A. Unique TensorFlower 提交于
PiperOrigin-RevId: 262650432
-
由 Yanhua Sun 提交于
PiperOrigin-RevId: 262649875
-
由 George Karpenkov 提交于
PiperOrigin-RevId: 262649330
-
由 Ayush Dubey 提交于
PiperOrigin-RevId: 262647780
-
由 TensorFlower Gardener 提交于
PiperOrigin-RevId: 262643291
-
由 A. Unique TensorFlower 提交于
Avoid diamond inheritance in tablegen with classes that inherit from Op and also inherit from Results and Arguments. Remove extraneous Operand and Result names in cases where there's only one. There will already be a generated getOperand/getResult method. Having an additional "res" accessor doesn't add value. In places where naming a single result is necessary for traits that check directly against attribute names, named single results "result". Avoid non-elementwise ops inheriting from elementwise subclasses. PiperOrigin-RevId: 262642541
-
https://bitbucket.org/eigen/eigen/commits/8071cda5714d7f454205b60b2f851986d53e8f33由 A. Unique TensorFlower 提交于
PiperOrigin-RevId: 262642073
-
由 Yifei Feng 提交于
Switch back to use easy_install to install pip2 and pip3, but let pip be upgraded to the latest by not specifying the version. PiperOrigin-RevId: 262639444
-
由 Edward Loper 提交于
Update the TensorInfo protobuf message with an encoding for composite tensors; and update SavedModel to use this new encoding. PiperOrigin-RevId: 262639435
-
由 TensorFlower Gardener 提交于
PiperOrigin-RevId: 262639287
-
由 A. Unique TensorFlower 提交于
PiperOrigin-RevId: 262639009
-
由 Derek Murray 提交于
Currently, GrpcServer does not support clean shutdown and destruction after GrpcServer::Start() is called, and logs a FATAL error in this case. In UpdateTFE_ContextWithServerDef(), there are several cases where a method can fail (e.g. due to invalid user input), and we attempt to destroy the created GrpcServer before returning. By deferring the call to GrpcServer::Start(), we can return an error to the client without crashing their process. PiperOrigin-RevId: 262638388
-
由 Yanhua Sun 提交于
export enable_tensor_equality disable_tensor_equality so that users have way to opt-in and opt-out explicitly PiperOrigin-RevId: 262637047
-
由 Gunhan Gulsoy 提交于
PiperOrigin-RevId: 262631581
-
由 Katherine Wu 提交于
PiperOrigin-RevId: 262630016
-
由 Benoit Jacob 提交于
PiperOrigin-RevId: 262628782
-
由 Scott Zhu 提交于
This will let the training function to return the correct number of examples to the callbacks. PiperOrigin-RevId: 262623679
-
由 Haoliang Zhang 提交于
PiperOrigin-RevId: 262622407
-
由 Ayush Dubey 提交于
PiperOrigin-RevId: 262615881
-
由 Alex Stark 提交于
PiperOrigin-RevId: 262613092
-
由 Reed Wanderman-Milne 提交于
These tests are both failing for unknown reasons on Windows, so we are disabling them until we can figure out the issue. PiperOrigin-RevId: 262610531
-
由 A. Unique TensorFlower 提交于
PiperOrigin-RevId: 262603945
-
由 Reed Wanderman-Milne 提交于
PiperOrigin-RevId: 262603680
-
由 Reed Wanderman-Milne 提交于
I fixed the examples so they actually can run now. And I mentioned that currently, only the first argument to call() is casted. PiperOrigin-RevId: 262603558
-
由 Yu-Cheng Ling 提交于
PiperOrigin-RevId: 262603535
-
由 Jing Pu 提交于
This also fix a crash in `condFn.getType()` if the lookupSymbol fails. PiperOrigin-RevId: 262599229
-
由 Feng Liu 提交于
We have observed that the tfl.mul folder couldn't fold the two operands even they are broadcast-compatible. Since we can guarantee the tf.Mul kernel can fold this, we switch to use tf.Mul instead. PiperOrigin-RevId: 262598998
-