| Commit message (Collapse) | Author | Age |
... | |
|
|
|
| |
PiperOrigin-RevId: 162289810
|
|
|
|
| |
PiperOrigin-RevId: 162284275
|
|
|
|
| |
PiperOrigin-RevId: 162282906
|
|
|
|
| |
PiperOrigin-RevId: 162277054
|
|
|
|
|
|
|
|
|
| |
END_PUBLIC
BEGIN_PUBLIC
Automated g4 rollback of changelist 162249449
PiperOrigin-RevId: 162276286
|
|
|
|
| |
PiperOrigin-RevId: 162275084
|
|
|
|
|
|
| |
include decision_trees in pip package properly.
PiperOrigin-RevId: 162273792
|
|
|
|
|
|
| |
tensorflow::Flag.
PiperOrigin-RevId: 162271241
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This changes the way the CPU backend emits slice operations
so that the maximal contiguous blocks get copied with memcpy.
Without this change, LLVM's optimizers prefer to unroll the
elementwise loops emitted for doing slice. This leads to increased
code size. With this change, the inliner only inlines the memcpy
when it is small enough.
This change also adds a test that exercises the logic to find
maximal contiguous blocks.
PiperOrigin-RevId: 162249449
|
|
|
|
|
|
|
| |
This contains a gRPC client that starts/stops tracing and processes/stores the result trace data into a TensorBoard log directory.
This also exposes trace_events proto classes via tf.contrib.tpu.profiler public API so that TensorBoard's profile plugin can process and visualize the profile.
PiperOrigin-RevId: 162247333
|
|
|
|
| |
PiperOrigin-RevId: 162244864
|
|
|
|
|
|
|
|
|
| |
In addition, this change:
- Adds a SWIG typemap to automatically unwrap ScopedTFStatus's for TF_Status* inputs
- Creates a new c_api_util.py file
- Switches a few TF_Status users to use raise_exception_on_not_ok_status() instead
PiperOrigin-RevId: 162243205
|
|
|
|
| |
PiperOrigin-RevId: 162230001
|
|
|
|
| |
PiperOrigin-RevId: 162222071
|
|
|
|
|
|
|
|
| |
This change also separates shared functionality into
c_test_util.h/cc. This brings c_api_test.cc to a mere 1715 LOC
(further splits can be more easily done now too).
PiperOrigin-RevId: 162216399
|
|
|
|
|
|
| |
instead of depending on variable initialization logic.
PiperOrigin-RevId: 162215450
|
|
|
|
| |
PiperOrigin-RevId: 162078550
|
|
|
|
|
|
| |
maximum size (in total bytes of cached block contents). The default maximum cache size is equal to the block size, and can be overridden by the environment variable GCS_READ_CACHE_MAX_SIZE_MB. In addition, deleting files now flushes them from the block cache, and renaming files flushes both the source and destination files from the block cache.
PiperOrigin-RevId: 162041685
|
|
|
|
| |
PiperOrigin-RevId: 162040682
|
|
|
|
| |
PiperOrigin-RevId: 162034086
|
|
|
|
| |
PiperOrigin-RevId: 162030530
|
|
|
|
|
|
|
| |
- http://stackoverflow.com/q/45109305
- #10766
PiperOrigin-RevId: 162026912
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
/device:GPU:0 compares to /gpu:0 as one would expect.
Adds a test to ensure that the warning message isn't printed
by checking the resulting device name. No behavior change
is expected here; all this does is change to share
the canonical name and remove the warning.
Fixes #11484.
PiperOrigin-RevId: 162024152
|
|
|
|
| |
PiperOrigin-RevId: 162022981
|
|
|
|
| |
PiperOrigin-RevId: 162022193
|
|
|
|
|
|
| |
This is the inverse operation of tf.contrib.signal.frame. This is implemented using GPU-capable ops and supports gradients.
PiperOrigin-RevId: 162017464
|
|
|
|
|
|
| |
erroneously claimed that scalar arguments should be vectors.
PiperOrigin-RevId: 162009915
|
|
|
|
| |
PiperOrigin-RevId: 162008248
|
|
|
|
| |
PiperOrigin-RevId: 162002161
|
|
|
|
|
|
|
|
|
|
|
| |
- Support arbitrary rank tensors.
- Add an `axis` parameter for framing any axis.
- Add a `pad_end` and `pad_value` parameter for controlling the signal padding. Padding is disabled by default to avoid a copy of a potentially large Tensor.
- Support shape inference of resulting framed tensor.
- Expand the tests, including tests for the gradients.
- Since `frames` is a noun and not a verb, rename `frames` to `frame`.
PiperOrigin-RevId: 161998921
|
|
|
|
| |
PiperOrigin-RevId: 161997142
|
|
|
|
|
|
|
|
|
|
| |
consolidate R1 tests.
* Fix SliceUnaligned1024In4096Values, which was using
unitialized data.
* Also fix SliceR2Test to respect the layout parameter.
PiperOrigin-RevId: 161991313
|
|
|
|
| |
PiperOrigin-RevId: 161990256
|
|
|
|
| |
PiperOrigin-RevId: 161981809
|
|
|
|
| |
PiperOrigin-RevId: 161979883
|
|
|
|
| |
PiperOrigin-RevId: 161972749
|
|
|
|
| |
PiperOrigin-RevId: 161967860
|
|
|
|
| |
PiperOrigin-RevId: 161967071
|
|
|
|
| |
PiperOrigin-RevId: 161949540
|
|
|
|
|
|
| |
It still tests wrap-around behavior, but in the most minor dimension instead of in the second most minor dimension.
PiperOrigin-RevId: 161924370
|
|
|
|
|
|
| |
guide.
PiperOrigin-RevId: 161917422
|
|
|
|
| |
PiperOrigin-RevId: 161906344
|
|
|
|
| |
PiperOrigin-RevId: 161905196
|
|
|
|
| |
PiperOrigin-RevId: 161903640
|
|
|
|
|
|
| |
file contents. Maximum staleness is specified in seconds in the GCS_READ_CACHE_MAX_STALENESS environment variable, and defaults to 0 (indicating that staleness is not tolerated in a newly opened file). Staleness is measured from the arrival time of the first block in the file's block cache.
PiperOrigin-RevId: 161897769
|
|
|
|
| |
PiperOrigin-RevId: 161897268
|
|
|
|
| |
PiperOrigin-RevId: 161896495
|
|
|
|
| |
PiperOrigin-RevId: 161896415
|
|
|
|
|
|
|
|
|
|
|
| |
Previously the CPU runtime wouldn't know how the data that was being outfed was
laid out by the XLA LayoutAssignment pass, which could result in
transposed-value results.
This also allows us to validate the contract between the host program and the
compiled XLA program with (reified) runtime type checks.
PiperOrigin-RevId: 161895093
|
|
|
|
|
|
| |
- TextLineDataset now accepts the compression_type argument.
PiperOrigin-RevId: 161892408
|