| COMMIT |
0.30 |
Automated Code Change |
|
Impersonal, formulaic phrase 'Automated |
2026-03-30 |
| PR |
0.10 |
Replace static recomputation op list with dynamic cost model |
|
Technical summary, domain references, no |
2026-02-02 |
| PR |
0.10 |
Fix tf.dynamic_stitch gradient for duplicate indices |
|
Technical details and informal, bullet s |
2026-02-02 |
| PR |
0.10 |
Fix integer overflow in stablehlo_pad output dimension compu |
|
Detailed vulnerability description, huma |
2026-03-29 |
| PR |
0.10 |
Validate window_size and stride in AudioSpectrogramOp constr |
|
Technical repro, jargon, concise style; |
2026-02-01 |
| PR |
0.10 |
Replace CHECK_GE with OP_REQUIRES in ReductionOp::Compute |
|
Technical, uses repro code and domain ja |
2026-02-01 |
| PR |
0.10 |
Bump Compute Library version from v24.12 to v52.8.0. |
|
Succinct update message, standard change |
2026-03-26 |
| COMMIT |
0.00 |
Translate async collectives between StableHLO and HLO. |
|
Brief technical phrase with project-spec |
2026-03-30 |
| COMMIT |
0.00 |
Update DynamicUpdateSlice for safer bytes size in memcpy. |
|
Short, technical update; human style, no |
2026-03-30 |
| COMMIT |
0.00 |
Introduce DeviceInterconnectResource to bridge Global Topolo |
|
Technical explanation, some structure bu |
2026-03-30 |
| COMMIT |
0.00 |
Refactor `CommonPjRtClient::DefineBuffer` to take `shared_pt |
|
Natural technical changelog, informal an |
2026-03-30 |
| COMMIT |
0.00 |
Reverts e117560c18eb072d8f41622f625acb18d8e4d48d |
|
Standard revert message with repo conven |
2026-03-30 |
| COMMIT |
0.00 |
[StreamExecutor] RAII-compliant Resource attachment via Pimp |
|
Technical language, example code; minor |
2026-03-30 |
| COMMIT |
0.00 |
PR #40112: [xla:gpu] Correctly report thunk execution index |
|
Detailed but informal, domain-specific t |
2026-03-30 |
| COMMIT |
0.00 |
PR #40053: [xla:ifrt] Allow passing PjRt client for autotuni |
|
Technical explanation, domain terms, hum |
2026-03-30 |
| COMMIT |
0.00 |
Integrate Triton up to [565c08520d](https://github.com/trito |
|
Concise integration message with commit |
2026-03-30 |
| COMMIT |
0.00 |
Update XNNPACK in tensorflow/xla |
|
Commit log with domain-specific referenc |
2026-03-30 |
| COMMIT |
0.00 |
Automated Code Change |
|
Generic short commit; no AI hallmarks. |
2026-03-30 |
| COMMIT |
0.00 |
[XLA:GPU] Add pad via the new tiling. |
|
Brief technical message; domain-specific |
2026-03-30 |
| COMMIT |
0.00 |
Remove TensorBoard dependency from TensorFlow |
|
Direct human-like commit message, terse |
2026-03-30 |
| COMMIT |
0.00 |
Reverts a0a1f5b376ab08c50c6559c4d43804e9d3634e27 |
|
Standard revert commit format; human-nor |
2026-03-30 |
| COMMIT |
0.00 |
Clean up coordination service `client_server_test.cc`. |
|
Bullet points, informal tone, and techni |
2026-03-30 |
| COMMIT |
0.00 |
Add a `device_memory_limit_bytes` attribute to PjRt GPU devi |
|
Succinct commit with technical detail, t |
2026-03-30 |
| COMMIT |
0.00 |
Support running IFRT through TFRT session on tf serving mode |
|
Technical phrasing and abbreviations; no |
2026-03-30 |
| COMMIT |
0.00 |
PR #40017: [xla:gpu] Correctly track thunk progress in prese |
|
Slightly more formal and elaborated, but |
2026-03-30 |
| COMMIT |
0.00 |
Add reshards for when func has no result shardings on shardy |
|
Dense technical phrasing, human-typical |
2026-03-30 |
| COMMIT |
0.00 |
[SC Serving] Bypass `sample_ids` scatter for univalent featu |
|
Short, domain-specific summary; nothing |
2026-03-30 |
| COMMIT |
0.00 |
Remove timeout long for slice_test. |
|
Concise, uses domain terms, no AI hallma |
2026-03-30 |
| COMMIT |
0.00 |
Log configs in coordination service `client_server_test.cc` |
|
Terse, informal, technical context, no A |
2026-03-30 |
| COMMIT |
0.00 |
Support extracting multiple consumers and use that in autotu |
|
Brief, technical, uses domain terminolog |
2026-03-30 |
| COMMIT |
0.00 |
[xla] Make some FloatNormalization tests more readable. |
|
Domain-specific, bullet points, informal |
2026-03-30 |
| COMMIT |
0.00 |
Migrate XLA GPU emitters and HWIR to use SymbolicMap. |
|
Technical content listing, informal phra |
2026-03-30 |
| COMMIT |
0.00 |
Add reshards for when func has no result shardings on unflat |
|
Concise, domain-specific, no AI markers. |
2026-03-30 |
| COMMIT |
0.00 |
disable flaky test |
|
Extremely terse, typical human flag of f |
2026-03-30 |
| COMMIT |
0.00 |
Rename `XNNPACK_CACHE_NO_MMAP_FOR_TEST` to `XNNPACK_CACHE_NO |
|
Informal with direct rationale, technica |
2026-03-30 |
| COMMIT |
0.00 |
Add dummy execute handler for CubScanUnassignedScratchSizeTa |
|
Technical and detailed, but some sentenc |
2026-03-30 |
| COMMIT |
0.00 |
[XLA:GPU] Preserve point dimension for concat if it occurs i |
|
Succinct technical commit message; no AI |
2026-03-30 |
| COMMIT |
0.00 |
Make CPU devices report as cpu:0 instead of TFRT_CPU_0. |
|
Brief, domain-specific phrasing; human s |
2026-03-30 |
| COMMIT |
0.00 |
[XLA:GPU][NFC] Pad the tile sizes before the tiling propagat |
|
Jargon and terse formatting; typical hum |
2026-03-30 |
| COMMIT |
0.00 |
[XLA:GPU] Reroute dots to a detailed cost model in the index |
|
Technical, concise; no AI stylistic indi |
2026-03-30 |
| COMMIT |
0.00 |
[XLA:GPU] Move ScatterDeterminismExpander pass to backends/g |
|
Direct explanation; informal tone; human |
2026-03-30 |
| COMMIT |
0.00 |
[xla] Add FloatNormalizationExcessPrecisionTest |
|
Technical summary, human abbreviations; |
2026-03-30 |
| COMMIT |
0.00 |
[XLA:GPU] Add transposes via the new tiling. |
|
Short, technical, domain-specific wordin |
2026-03-30 |
| COMMIT |
0.00 |
[XLA:GPU] Use copied device info to init members of GpuCostM |
|
Terse, technical explanation; human comm |
2026-03-30 |
| COMMIT |
0.00 |
PR #39950: [ROCm] Fix hipblasLt Int8 GEMM support and autotu |
|
PR template detected, but free-text is d |
2026-03-30 |
| COMMIT |
0.00 |
Disable unflattener for Shardy to GSPMD fallback. |
|
Technical phrasing, domain jargon; human |
2026-03-30 |
| COMMIT |
0.00 |
[XLA:GPU] Fix ReorderFilterAndBiasHloTest.TestCudnnReorderFi |
|
Technical explanation, precise, informal |
2026-03-30 |
| COMMIT |
0.00 |
[XLA:GPU] Add support for constants via new tiling. |
|
Extremely brief, minimal, typical human |
2026-03-30 |
| COMMIT |
0.00 |
Automated Code Change |
|
Standard automated commit info, no AI in |
2026-03-30 |
| COMMIT |
0.00 |
PR #39931: Add test coverage for AllGatherRemoveDegenerateDi |
|
Template with detailed technical test co |
2026-03-30 |
| COMMIT |
0.00 |
PR #39508: [XLA:GPU][oneAPI] Remove hardcoded spirv-binary i |
|
Technical, specific, direct human-writte |
2026-03-30 |
| COMMIT |
0.00 |
PR #39417: [ROCm] Implement empty graph node support and saf |
|
Technical summary, emoji use, detailed, |
2026-03-30 |
| COMMIT |
0.00 |
PR #39956: [xla:gpu] Log thunk progress in chronological ord |
|
Direct explanation of confusion and solu |
2026-03-30 |
| COMMIT |
0.00 |
Automated Code Change |
|
Standard automated commit info, no AI in |
2026-03-30 |
| COMMIT |
0.00 |
PR #39025: [xla] Fix potentially expensive spin in ObjectPoo |
|
Very technical, domain-specific, informa |
2026-03-30 |
| COMMIT |
0.00 |
Automated Code Change |
|
Standard automated commit info, no AI in |
2026-03-30 |
| COMMIT |
0.00 |
[XLA:GPU] Add command buffer scheduling mode `CONCURRENT_REG |
|
Concise technical language, human-author |
2026-03-30 |
| COMMIT |
0.00 |
PR #39497: [XLA:GPU][oneAPI] Register base oneCCL collective |
|
Template sections with brief, domain-spe |
2026-03-30 |
| COMMIT |
0.00 |
[XLA:GPU]: Reapply flag flip for one shot all-reduce after f |
|
Brief, domain-specific, typical human re |
2026-03-30 |
| COMMIT |
0.00 |
Move [[nodiscard]] attribute from definition to declaration. |
|
Terse explanation; clear human engineeri |
2026-03-30 |
| COMMIT |
0.00 |
[XLA:GPU] Migrate TmaAndLayoutParameterizedTritonEmitterTest |
|
Concise, typical commit style, domain-sp |
2026-03-30 |
| COMMIT |
0.00 |
Hoist HoistFusedBitcasts above GemmFusionSwapOperands - this |
|
Technical phrasing with informal tone; h |
2026-03-30 |
| COMMIT |
0.00 |
[XLA:GPU] Migrate more triton tests to lit. |
|
Concise commit subject; no AI traits det |
2026-03-29 |
| COMMIT |
0.00 |
Add a new collective permute cse pass |
|
Brief, direct commit message; entirely h |
2026-03-29 |
| COMMIT |
0.00 |
Reverts 15ddf2c954cb94d03ae883a95b54cb4ad4660f1e |
|
Standard revert syntax; no AI indicators |
2026-03-29 |
| COMMIT |
0.00 |
RecognizeReduce: fix bug with subtract combining |
|
Bugfix note with technical detail; typic |
2026-03-29 |
| COMMIT |
0.00 |
[XLA:GPU] Remove NewTilingEmitterParameterizedTest and use t |
|
Concise change description; no AI hallma |
2026-03-29 |
| COMMIT |
0.00 |
Reverts 599f11d91fb2966208907684966a012e5b288837 |
|
Standard revert commit message; human-ge |
2026-03-29 |
| COMMIT |
0.00 |
Rollback to not trigger fatal log on reshard through full re |
|
Revert with clear human-written justific |
2026-03-29 |
| COMMIT |
0.00 |
PR #39382: [xla:gpu] Migrate collective thunks to generic As |
|
Technical language and changelog style; |
2026-03-28 |
| COMMIT |
0.00 |
Rollback of drop deduplicating on outliner. |
|
Succinct, domain-specific rollback messa |
2026-03-28 |
| COMMIT |
0.00 |
Restore discarded attribute |
|
Contains domain-specific detail and conc |
2026-03-28 |
| COMMIT |
0.00 |
Drop full deduplication on shardy outliner. |
|
Technical, focused, and informal changel |
2026-03-28 |
| COMMIT |
0.00 |
[XLA:GPU] Rename multimem parameter property |
|
Short, standard rename commit message. |
2026-03-28 |
| COMMIT |
0.00 |
Automated Code Change |
|
Standard automated change notice; no tex |
2026-03-28 |
| COMMIT |
0.00 |
Reduction detection, make more aggressive for overlapping sl |
|
Short, domain-specific update; no AI sig |
2026-03-28 |
| COMMIT |
0.00 |
Hosting builds for pull requests check on GitHub - Windows |
|
Concise hosting change; no AI-specific m |
2026-03-28 |
| COMMIT |
0.00 |
Update ReshardFromPartialReplicateWithDynamicSlice for named |
|
Brief, technical message with domain con |
2026-03-28 |
| COMMIT |
0.00 |
[XLA:MSA] Allow conditional inputs with uses after the condi |
|
Very technical and specific to XLA:MSA; |
2026-03-28 |
| COMMIT |
0.00 |
Add bound check to TFLite graph loader |
|
Succinct commit message, no AI indicator |
2026-03-28 |
| COMMIT |
0.00 |
Support passing a creation mode for TSL `RecursivelyCreateDi |
|
Technical, concise phrasing, no AI trait |
2026-03-28 |
| COMMIT |
0.00 |
Enhance IfrtRestoreVariableOp to support non-variable tensor |
|
Direct technical improvement, no AI hall |
2026-03-28 |
| COMMIT |
0.00 |
Reverts 42282d37d0fb5a8ee88478c7ad02df14305c39ba |
|
Common terse revert message; fully human |
2026-03-27 |
| COMMIT |
0.00 |
Convert shardings from V3 to V2 in the dot handler and param |
|
Specific, technical language; not AI-gen |
2026-03-27 |
| COMMIT |
0.00 |
[XLA:GPU] Get information about NVLink domain size from devi |
|
Uses domain-specific terms and informal |
2026-03-27 |
| COMMIT |
0.00 |
emit BitonicSort as CustomKernelThunk |
|
Technical commit message; terse and spec |
2026-03-27 |
| COMMIT |
0.00 |
Fix typo in `tsl::monitoring::Buckets::Exponential` document |
|
Concise human-style commit, domain-speci |
2026-03-27 |
| COMMIT |
0.00 |
Remove unused `TransactionToken` and associated methods from |
|
Removal refactoring message; no AI signa |
2026-03-27 |
| COMMIT |
0.00 |
Add support for Named sharding format in ShardingFormatPicke |
|
Detailed technical PR, domain terms, no |
2026-03-27 |
| COMMIT |
0.00 |
Migrate tuple_test to PjRt runtime. |
|
Direct, technical test migration with mi |
2026-03-27 |
| COMMIT |
0.00 |
Fix `GetMeshAxesPartitionGroupsForReplication` for NamedShar |
|
Technical fix explanation, clear domain |
2026-03-27 |
| COMMIT |
0.00 |
[IFRT IR] Remove ifrt_mpmd_py bindings are they've been repl |
|
Informal commit, typo but clear domain k |
2026-03-27 |
| COMMIT |
0.00 |
Simplify PjRtDeviceEvent by moving event tracking to CommonP |
|
Terse commit; technical detail, not AI-l |
2026-03-27 |
| COMMIT |
0.00 |
[PjRt-IFRT] Remove spurious warning messages during executab |
|
Technical explanation; informative, not |
2026-03-27 |
| COMMIT |
0.00 |
Integrate LLVM at llvm/llvm-project@9a0b003dde83 |
|
Repository sync message, standard for LL |
2026-03-27 |
| COMMIT |
0.00 |
Roll forward with fix |
|
Very terse, standard revert commit messa |
2026-03-27 |
| COMMIT |
0.00 |
New recognize_reduce_window_test unnecessarily depends on de |
|
Concise, correct, and technical human st |
2026-03-27 |
| COMMIT |
0.00 |
round_trip_packed_literal_test should not depend on client_l |
|
Short message with project-specific term |
2026-03-27 |
| COMMIT |
0.00 |
Fix kRecv resource limit violation in top-down scheduling. |
|
Direct and technical; common human commi |
2026-03-27 |
| COMMIT |
0.00 |
Internal clean up. |
|
Extremely brief, typical human commit me |
2026-03-27 |
| COMMIT |
0.00 |
[XLA:MSA] Allow conditional outputs in alternate memory. |
|
Uses project and domain-specific abbrevi |
2026-03-27 |
| COMMIT |
0.00 |
[XLA:GPU]: Fix layout when doing a gathered load. |
|
Technical detail and informal list; natu |
2026-03-27 |
| COMMIT |
0.00 |
Rewrite matching python_repo requirement entries with local |
|
Detailed but context-specific and techni |
2026-03-27 |
| COMMIT |
0.00 |
[XLA:GPU] Emit copy instruction when Mosaic operation is use |
|
Specialized explanation, natural human w |
2026-03-27 |
| COMMIT |
0.00 |
Integrate LLVM at llvm/llvm-project@5a403150b468 |
|
Standard integration message; explicit r |
2026-03-27 |
| COMMIT |
0.00 |
Add `Label` to `//third_party/tensorflow/compiler/xla/tsl/ut |
|
Succinct, domain-specific, no AI signals |
2026-03-27 |
| PR |
0.00 |
Add HLO constraint propagation for test argument generation |
|
— |
2026-03-18 |
| PR |
0.00 |
Translate async collectives between StableHLO and HLO. |
|
— |
2026-03-19 |
| PR |
0.00 |
[ReplicaGroupV3] Cleanup replica group parsing in HLO parser |
|
— |
2026-03-30 |
| PR |
0.00 |
[PjRt-IFRT] Fill in parameter information in SerializedXlaEx |
|
— |
2026-03-27 |
| PR |
0.00 |
Fix building TF CUDA wheel by adding dependency on stub alia |
|
— |
2026-03-30 |
| PR |
0.00 |
[PjRt] Implement `PjRt(Loaded)Executable::GetParameterMemory |
|
— |
2026-03-27 |
| PR |
0.00 |
Use absl::HashOf for key_hash in LocalRendezvous. |
|
— |
2026-03-30 |
| PR |
0.00 |
Remove `kLeaveBarriersOnRecoverableAgentRestart` from coordi |
|
— |
2026-03-30 |
| PR |
0.00 |
Integrate LLVM at llvm/llvm-project@1a08f41281bf |
|
— |
2026-03-30 |
| PR |
0.00 |
[XLA] Remove obsolete shmap & sdy manual computation inliner |
|
— |
2026-03-30 |
| PR |
0.00 |
Update DynamicUpdateSlice for safer bytes size in memcpy. |
|
— |
2026-03-20 |
| PR |
0.00 |
Allow non-kPadUpPolicy for priority-aware batching when max_ |
|
— |
2026-03-30 |
| PR |
0.00 |
Add PRNG seed to PjRt ExecuteOptions and add runtime support |
|
— |
2026-03-26 |
| PR |
0.00 |
Introduce DeviceInterconnectResource to bridge Global Topolo |
|
— |
2026-03-26 |
| PR |
0.00 |
Migrate xla::ifrt::GetDeviceMemoryInBytes to `device_memory_ |
|
— |
2026-03-30 |
| PR |
0.00 |
Integrate LLVM at llvm/llvm-project@6ead6868e7ca |
|
— |
2026-03-30 |
| PR |
0.00 |
Integrate LLVM at llvm/llvm-project@6ead6868e7ca |
|
— |
2026-03-30 |
| PR |
0.00 |
[xla:cpu] Make small_while_loop_hoisting_pass_test use the s |
|
— |
2026-03-30 |
| PR |
0.00 |
add GB200 and GB300 to GetDeviceMemoryInBytes. |
|
— |
2026-03-30 |
| PR |
0.00 |
Add GetOrCreateRuntimeError to PJRT Megascale Extension |
|
— |
2026-03-26 |
| PR |
0.00 |
Add PID metadata to XPlanes and TraceMeConsumers to enable p |
|
— |
2025-12-09 |
| PR |
0.00 |
Add CUDA 13 presubmit job and update Python dependencies. |
|
— |
2026-03-30 |
| PR |
0.00 |
Benchmark for traceme_encode. |
|
— |
2026-03-30 |
| PR |
0.00 |
Support the XLA GPU compilation flags in Orbax |
|
— |
2026-03-23 |
| PR |
0.00 |
Refactor `CommonPjRtClient::DefineBuffer` to take `shared_pt |
|
— |
2026-03-29 |
| PR |
0.00 |
1. Fixed Missing Profiles: Enabled modern telemetry to captu |
|
— |
2026-03-30 |
| PR |
0.00 |
Reverts e117560c18eb072d8f41622f625acb18d8e4d48d |
|
— |
2026-03-30 |
| PR |
0.00 |
Make an explicit ref type for PjRtDeviceEvent to allow easil |
|
— |
2026-03-27 |
| PR |
0.00 |
[StreamExecutor] RAII-compliant Resource attachment via Pimp |
|
— |
2026-03-27 |
| PR |
0.00 |
PR #40112: [xla:gpu] Correctly report thunk execution index |
|
— |
2026-03-30 |
| PR |
0.00 |
Automated Code Change |
|
— |
2026-01-10 |
| PR |
0.00 |
PR #40053: [xla:ifrt] Allow passing PjRt client for autotuni |
|
— |
2026-03-30 |
| PR |
0.00 |
Integrate Triton up to [565c08520d](https://github.com/trito |
|
— |
2026-03-30 |
| PR |
0.00 |
Automated Code Change |
|
— |
2026-01-10 |
| PR |
0.00 |
Fix use of deprecated `xnn_define_static_constant_pad` |
|
— |
2026-03-30 |
| PR |
0.00 |
Update XNNPACK in tensorflow/xla |
|
— |
2026-03-30 |
| PR |
0.00 |
Populate xla_input_shapes for unpacked inputs in IfrtServing |
|
— |
2026-03-26 |
| PR |
0.00 |
Bump pygments from 2.19.1 to 2.20.0 in /ci/official/requirem |
|
— |
2026-03-30 |
| PR |
0.00 |
Automated Code Change |
|
— |
2026-03-18 |
| PR |
0.00 |
Bump pygments from 2.19.1 to 2.20.0 |
|
— |
2026-03-30 |
| PR |
0.00 |
PR #39929: [ROCm] Fix parse empty rocm distro links |
|
— |
2026-03-30 |
| PR |
0.00 |
PR #39925: [ROCm] Remove deprecated hipCtx_t and HIP context |
|
— |
2026-03-30 |
| PR |
0.00 |
Collect runtime environment variables in XLA profiler metada |
|
— |
2026-03-27 |
| PR |
0.00 |
[XLA:GPU] Add pad via the new tiling. |
|
— |
2026-03-30 |
| PR |
0.00 |
Automated Code Change |
|
— |
2026-01-10 |
| PR |
0.00 |
Remove TensorBoard dependency from TensorFlow |
|
— |
2026-03-26 |
| PR |
0.00 |
Automated Code Change |
|
— |
2026-01-10 |
| PR |
0.00 |
Reverts a0a1f5b376ab08c50c6559c4d43804e9d3634e27 |
|
— |
2026-03-30 |
| PR |
0.00 |
Clean up coordination service `client_server_test.cc`. |
|
— |
2026-03-30 |
| PR |
0.00 |
PR #39695: Fix xla aot compile gpu test |
|
— |
2026-03-30 |
| PR |
0.00 |
Add a `device_memory_limit_bytes` attribute to PjRt GPU devi |
|
— |
2026-03-27 |
| PR |
0.00 |
Fix resource misclassification for unsorted _read_only_resou |
|
Technical summary with domain specifics, |
2026-02-18 |
| PR |
0.00 |
added same size validation for var and accum_update in Apply |
|
Direct bugfix explanation; human technic |
2026-02-16 |
| PR |
0.00 |
Support running IFRT through TFRT session on tf serving mode |
|
— |
2026-03-23 |
| PR |
0.00 |
PR #40017: [xla:gpu] Correctly track thunk progress in prese |
|
— |
2026-03-30 |
| PR |
0.00 |
Cleanup/ops todo deprecation removal |
|
Highly terse and informal, no AI signals |
2026-02-11 |
| PR |
0.00 |
Switch TSL Monitoring factory call sites from `T::New` to `T |
|
— |
2026-03-27 |
| PR |
0.00 |
Automated Code Change |
|
— |
2026-01-10 |
| PR |
0.00 |
Add reshards on func results also when func has no result sh |
|
— |
2026-03-26 |
| PR |
0.00 |
Rename TSL Monitoring's `MakeStatic` factory functions to `N |
|
— |
2026-03-26 |
| PR |
0.00 |
Automated Code Change |
|
— |
2026-01-10 |
| PR |
0.00 |
[SC Serving] Bypass `sample_ids` scatter for univalent featu |
|
— |
2026-03-27 |
| PR |
0.00 |
Remove timeout long for slice_test. |
|
— |
2026-03-30 |
| PR |
0.00 |
Integrate LLVM at llvm/llvm-project@5688aca96ea2 |
|
— |
2026-03-30 |
| PR |
0.00 |
Integrate LLVM at llvm/llvm-project@5688aca96ea2 |
|
— |
2026-03-30 |
| PR |
0.00 |
Add `stablehlo.future` to C API. |
|
— |
2026-03-30 |
| PR |
0.00 |
[SC Serving] Clamps embedding IDs strictly to `[0, vocab_siz |
|
— |
2026-03-27 |
| PR |
0.00 |
Automated Code Change |
|
— |
2026-01-10 |
| PR |
0.00 |
Log configs in coordination service `client_server_test.cc` |
|
— |
2026-03-25 |
| PR |
0.00 |
[XLA:GPU]: Unify collective codegen checks in one place. |
|
— |
2026-03-30 |
| PR |
0.00 |
Automated Code Change |
|
— |
2026-03-07 |
| PR |
0.00 |
Support extracting multiple consumers and use that in autotu |
|
— |
2026-03-30 |
| PR |
0.00 |
[xla] Make some FloatNormalization tests more readable. |
|
— |
2026-03-30 |
| PR |
0.00 |
PR #39871: [ROCm] Fix bf16 upcast handling for libdevice cal |
|
— |
2026-03-27 |
| PR |
0.00 |
[XLA:GPU] Added concat to the allowlist of block-level emitt |
|
— |
2026-03-30 |
| PR |
0.00 |
PR #39951: [xla:gpu] Add missing scheduling for all-gather-s |
|
— |
2026-03-26 |
| PR |
0.00 |
Migrate XLA GPU emitters and HWIR to use SymbolicMap. |
|
— |
2026-02-10 |
| PR |
0.00 |
Integrate LLVM at llvm/llvm-project@fc01c81d03c7 |
|
— |
2026-03-30 |
| PR |
0.00 |
Add reshards for when func has no result shardings on unflat |
|
— |
2026-03-26 |
| PR |
0.00 |
disable flaky test |
|
— |
2026-03-30 |
| PR |
0.00 |
Move in/outliner xla utility into shardy. |
|
— |
2026-03-30 |
| PR |
0.00 |
Automated Code Change |
|
— |
2026-01-10 |
| PR |
0.00 |
Rename `XNNPACK_CACHE_NO_MMAP_FOR_TEST` to `XNNPACK_CACHE_NO |
|
— |
2026-03-27 |
| PR |
0.00 |
Remove legacy autotuner compile util. |
|
— |
2026-03-26 |
| PR |
0.00 |
Add dummy execute handler for CubScanUnassignedScratchSizeTa |
|
— |
2026-03-30 |
| PR |
0.00 |
Fully serialize stream_executor::DeviceDescription to/from p |
|
— |
2026-03-30 |
| PR |
0.00 |
[XLA:GPU] Preserve point dimension for concat if it occurs i |
|
— |
2026-03-30 |
| PR |
0.00 |
Make CPU devices report as cpu:0 instead of TFRT_CPU_0. |
|
— |
2026-03-27 |
| PR |
0.00 |
Always populate target config in GpuTopology created in PjRt |
|
— |
2026-03-26 |
| PR |
0.00 |
Remove redundant dot_operation_single_threaded_runtime_test. |
|
— |
2026-03-30 |
| PR |
0.00 |
[XLA:GPU][NFC] Pad the tile sizes before the tiling propagat |
|
— |
2026-03-30 |
| PR |
0.00 |
[XLA:GPU] Prevent collective memory from being destructed un |
|
— |
2026-03-30 |
| PR |
0.00 |
[XLA:GPU] Reroute dots to a detailed cost model in the index |
|
— |
2026-03-26 |
| PR |
0.00 |
Automated Code Change |
|
— |
2026-01-10 |
