Merging updates from Tensorflow upstream master - 07/09/2018 #62
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
PiperOrigin-RevId: 203557079
PiperOrigin-RevId: 203561889
When storing images in Cloud Bigtable, the resulting gRPC messages are often larger than the default receive message max size value. This change makes the maximum receive message sizes configurable, and sets a more reasonable default for general TensorFlow use. PiperOrigin-RevId: 203569796
Restructured sharding passes to propagate sharding on pass-through instructions which now the placer does not assign anymore (GTEs, tuples, bitcast, parameters, ...). PiperOrigin-RevId: 203591020
configure.py respects CUDA_TOOLKIT_PATH instead of CUDA_INSTALL_PATH PiperOrigin-RevId: 203591214
…ite Object Detection app. PiperOrigin-RevId: 203689941
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=55971 PiperOrigin-RevId: 203698458
PiperOrigin-RevId: 203698572
PiperOrigin-RevId: 203708965
Instead of having one stream for compute, host-to-device and device-to-host transfers, switch to having separate streams, just like the GPU does. Add a se::Event field to XlaTensor to allow accurate inter-stream dependencies to be created. As part of this: - Fix TransferManager::TransferLiteralFrom/ToDevice to correctly make generated substreams wait on their master stream. - Fix Stream::BlockHostUntilDone() to not block on or return substreams. This behavior is completely broken and not only nondeterministically returns substreams to the pool but causes indefinite hangs with the HostStream. PiperOrigin-RevId: 203726543
When doing multi-output fusion and using sibling fusion, it can happen that we don't need to clone the 'instruction_to_fuse' argument. Right now, we clone, and then delete the clone again, and at the end of the function try to print the debug string for the clone (which then crashes). Instead, we can simply not generate the clone if it is not needed, and catch this case before printing the debug string. PiperOrigin-RevId: 203733796
If a domain become empty because the various optimizations removed all instruction from it then we have to re-add some instruction to make sure the user supplied sharding is still respected. This is especially important for the root instruction as the user will expect the data to be available on the device they requested it. Before this CL we failed to insert the tuple->gte sequence into the empty domain due to a bug where we only considered cases where we have an exit domain what is not the case for the root instruction. PiperOrigin-RevId: 203744534
Benchmark should emit info even if extras is None. PiperOrigin-RevId: 203762356
PiperOrigin-RevId: 203763201
PiperOrigin-RevId: 203766656
PiperOrigin-RevId: 203768649
PiperOrigin-RevId: 203769116
PiperOrigin-RevId: 203773485
Collaborator
|
@deven-amd please help review those failures and update whitelist if necessary. |
Author
|
only two tests are failing.
I cannot reproduce the failure for #1 locally (3 consecutuive passes for me) Failure #2 is reproducible (crash with stack dump while running) but currently do not know the cause of of the failure/crash. |
Collaborator
|
a) let's put them into the whitelist, and update the PR |
Author
|
let me put the two tests on the whitelist and update the PR The dnn_linear_combined_test is a regression (though I cannot reproduce the failure locally) ... it started passing recently (either due to disabling sharding or the "FloorDiv on GPU" fix) The parallel_for/control_ops_test is a new test! |
//tensorflow/python/estimator:dnn_linear_combined_test is a regression. However I cannot reproduce the failure locally //tensorflow/python/ops/parallel_for:control_flow_ops_test is a new test. There is also another "control_flow_ops" test on the whitelist...perhaps a common cause of failure.
deven-amd
pushed a commit
that referenced
this pull request
Aug 2, 2019
Closes #62 COPYBARA_INTEGRATE_REVIEW=tensorflow/mlir#62 from schweitzpgi:register-fir d122eae9c2cdf21581f48412551a93b8b4e640a6 PiperOrigin-RevId: 261187850
jerryyin
pushed a commit
that referenced
this pull request
Oct 3, 2019
Closes #62 PiperOrigin-RevId: 261187850
ekuznetsov139
pushed a commit
that referenced
this pull request
May 31, 2022
PiperOrigin-RevId: 449102807
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.