forked from tensorflow/tensorflow
-
Notifications
You must be signed in to change notification settings - Fork 95
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Develop upstream sync 241105 #2744
Merged
Merged
Conversation
This file contains 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: 691456562
PiperOrigin-RevId: 691468873
This can be reverted to getCurrentVersion once frameworks release with the fix in: openxla/xla@2f99455 Until then, a plugin that is newer than its framework will error on serialization, this feature was added in 1.7.X, so pinning to 1.7.0 should be safe. PiperOrigin-RevId: 691471627
PiperOrigin-RevId: 691480719
…-with-null-data-for-simple-dynamic-buffer PiperOrigin-RevId: 691481168
…ling test cases. Only test cases breaking on CPU are related to: - pure callbacks - export - shard alike Note that `layout_test` is broken on TPU, leaving a comment saying to enable it. Also fixed `shard_map_test` test that was broken when running Shardy on one TPU, and `aot_test` which was breaking due to calling a different C++ StableHLO compilation function. PiperOrigin-RevId: 691496997
…to stay consistent with hermetic CUDA PiperOrigin-RevId: 691506371
PiperOrigin-RevId: 691516394
…low/compiler/xla/service/spmd/shardy/mhlo_round_trip/export_shardings.cc. PiperOrigin-RevId: 691525161
PiperOrigin-RevId: 691528885
PiperOrigin-RevId: 691565219
2. Set tasks to error (and don't disconnect tasks) during a failed shutdown to avoid silent reconnects) PiperOrigin-RevId: 691586482
Updates LLVM usage to match [4ba623f24479](llvm/llvm-project@4ba623f24479) PiperOrigin-RevId: 691589403
…til.h/cc PiperOrigin-RevId: 691591073
PiperOrigin-RevId: 691602092
PiperOrigin-RevId: 691648381
And fix version check for Dispatch API PiperOrigin-RevId: 691663743
PiperOrigin-RevId: 691667184
PiperOrigin-RevId: 691667192
PiperOrigin-RevId: 691675947
PiperOrigin-RevId: 691681901
PiperOrigin-RevId: 691692633
PiperOrigin-RevId: 691718921
PiperOrigin-RevId: 691720760
PiperOrigin-RevId: 691720775
PiperOrigin-RevId: 691739227
…introduced in `c3d5769` Imported from GitHub PR openxla/xla#18860 Copybara import of the project: -- 1b61efe8270e67140199bbbb70665955fbaa6656 by Harsha HS <Harsha.HavanurShamsundara@amd.com>: [ROCm] Remove IsEmpty check for execution_order introduced in c3d5769 Merging this change closes tensorflow#18860 PiperOrigin-RevId: 691744865
These functions are not declared in any header file, hence they should have internal linkage. PiperOrigin-RevId: 691746918
…legacy emitters. The default lowering in the MLIR repo is not stable for small imag(arg). PiperOrigin-RevId: 693262812
PiperOrigin-RevId: 693264535
… efficently PiperOrigin-RevId: 693265606
The test failures are due to the fact that the names of kernels from CUDA are not deterministic. PiperOrigin-RevId: 693272080
…il out if it fails. Add a large number of tests extracted from triton_fusion_emitter_device_legacy_test. PiperOrigin-RevId: 693278160
PiperOrigin-RevId: 693279863
PiperOrigin-RevId: 693283395
…ir/tensorflow/transforms/executor_island_coarsening.cc PiperOrigin-RevId: 693288112
PiperOrigin-RevId: 693289401
Imported from GitHub PR openxla/xla#18948 Copybara import of the project: -- 80e717c39e8a120cca974dca9f473d817d3a3457 by Ilia Sergachev <isergachev@nvidia.com>: [GPU][NFC] Improve error messages. Merging this change closes tensorflow#18948 PiperOrigin-RevId: 693291127
…/compiler/mlir/tosa/transforms/convert_tfl_uint8.cc. PiperOrigin-RevId: 693300221
…s much as possible. This is particularly useful in FSDP/HSDP where gradient propagation can be done fully in the i+1th iteration. It takes the responsibility of the user to set the `xla_gpu_all_reduce_combine_threshold_bytes` by themselves. PiperOrigin-RevId: 693304915
retest gpu-pycpp please |
This reverts commit 1c5d7d4.
hsharsha
approved these changes
Nov 29, 2024
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.