[go: nahoru, domu]
Skip to content

Pull requests: tensorflow/tensorflow

New pull request New
1,351 Open 29,265 Closed
1,351 Open 29,265 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

[tsl] Mark tsl::NullTerminatedMessage() with ABSL_DEPRECATE_AND_INLINE
#64435 opened Mar 25, 2024 by copybara-service bot • Draft
[tsl] Implement tsl::NullTerminatedMessage() in terms of absl::StatusMessageAsCStr()
#64433 opened Mar 25, 2024 by copybara-service bot • Draft
EXPERIMENT: change RestoreV2 to PureOp
#64427 opened Mar 25, 2024 by copybara-service bot • Draft
To ensure that C++20 is fully supported in TensorFlow,
#64424 opened Mar 25, 2024 by copybara-service bot • Draft
Use hermetic Python in JAX. See requirements_updater/README.md for details.
#64395 opened Mar 25, 2024 by copybara-service bot • Draft
Use hermetic Python in TSL and XLA
#64394 opened Mar 25, 2024 by copybara-service bot • Draft
MakeArrayFromHostBuffer and AssembleArrayFromSingleDeviceArrays can be shared with all PjRtCompatibleClient subclasses. No sense making them specialized to only work for PjRtClients.
#64340 opened Mar 23, 2024 by copybara-service bot • Draft
No need to include all of Eigen Tensor to get the threadpool after we moved it to Core.
#64337 opened Mar 22, 2024 by copybara-service bot • Draft
Create streams using StreamExecutor::CreateStream rather than creating them and calling Initialize.
#64335 opened Mar 22, 2024 by copybara-service bot • Draft
[xla:pjrt:cpu] Create owned PjRt buffers when creating zero-copy host buffers
#64330 opened Mar 22, 2024 by copybara-service bot • Draft
[XLA] Add to ReduceScatterDecomposer target specific hook to allow transforming the reduce-scatter to a supported version
#64312 opened Mar 22, 2024 by copybara-service bot • Draft
[XLA] Prioritize instructions that are on mostly a latency path (not a lot of compute on the path) as they can be overlapped with all the compute available.
#64309 opened Mar 22, 2024 by copybara-service bot • Draft
Add a new field inference into streamz instrument UpdateTfMlirBridgeFirstPhaseCounter
#64300 opened Mar 22, 2024 by copybara-service bot • Draft
Always enable XLA spmd
#64175 opened Mar 21, 2024 by copybara-service bot • Draft
Introduce hermetic CUDA in Google ML projects.
#64130 opened Mar 21, 2024 by copybara-service bot • Draft
1
Add int4 support to most remaining ops on CPU/GPU.
#64110 opened Mar 21, 2024 by copybara-service bot • Draft
Internal config test
#64108 opened Mar 21, 2024 by copybara-service bot • Draft
Test 1
#64106 opened Mar 21, 2024 by copybara-service bot • Draft
Migrate to new header
#64084 opened Mar 20, 2024 by copybara-service bot • Draft
Clean up some deps in flatbuffer_import, flatbuffer_export.
#64074 opened Mar 20, 2024 by copybara-service bot • Draft
Update docs to make use of new API for adding a TfLiteRegistrationExternal to a MutableOpResolver.
#64062 opened Mar 20, 2024 by copybara-service bot • Draft
PR #10668: [XLA:GPU] Fix cuDNN FMHA rewriter sequence length checks and Bump up minimum flash attn cuDNN version to 8.9.4
#64044 opened Mar 20, 2024 by copybara-service bot • Draft
1
Fixed typos in post_training_quant.ipynb awaiting review Pull request awaiting review comp:lite TF Lite related issues size:S CL Change Size: Small
#64028 opened Mar 20, 2024 by LakshmiKalaKadali Loading…
@gbaned
4
Internal change only
#64018 opened Mar 20, 2024 by copybara-service bot • Draft
Allow MaybeSavedModelDirectory to read via custom env.
#64002 opened Mar 19, 2024 by copybara-service bot • Draft
ProTip! Exclude everything labeled bug with -label:bug.