Skip to content

Pull requests: openxla/xla

New pull request New
1,885 Open 16,008 Closed
1,885 Open 16,008 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

Move stream handle from GpuStream to CudaStream and RocmStream
#18306 opened Oct 14, 2024 by copybara-service bot Loading…
@beckerhe
Integrate LLVM at llvm/llvm-project@1a787b3c8e71
#18305 opened Oct 14, 2024 by copybara-service bot Loading…
[XLA] Add compilation effort flags.
#18304 opened Oct 14, 2024 by copybara-service bot Loading…
Use ModuleHandle for module state tracking
#18303 opened Oct 14, 2024 by copybara-service bot Loading…
@beckerhe
Add tests for CudaStream::WaitFor
#18302 opened Oct 14, 2024 by copybara-service bot Loading…
@beckerhe
Remove LeakCheckDisabler from CudaExecutor
#18301 opened Oct 14, 2024 by copybara-service bot Loading…
@beckerhe
More graceful handling AbslHashValue when hlo module doesn't have entry computation layout. Without the fix hash breaks on empty module such as absl:Hash(HloTestBase::CreateNewVerifiedModule()).
#18300 opened Oct 14, 2024 by copybara-service bot Loading…
@toli-y
Use StreamExecutor::Activate in ROCm code.
#18299 opened Oct 14, 2024 by copybara-service bot Loading…
@klucke
Add AbslStringify to most "simple" IFRT types
#18298 opened Oct 14, 2024 by copybara-service bot Loading…
PR #18254: [XLA:GPU] Simplifies All-Reduce if they can be simplified or are degenerated.
#18297 opened Oct 14, 2024 by copybara-service bot Loading…
[Rocm] fix arch
#18296 opened Oct 14, 2024 by Ruturaj4 Loading…
1
Use StreamExecutor::Activate instead of directly creating ScopedActivateContexts in cuda code.
#18294 opened Oct 14, 2024 by copybara-service bot Loading…
@klucke
[XLA:GPU] Large dynamic update slice fusion due to inability to identify in place fusion
#18292 opened Oct 14, 2024 by Cjkkkk Loading…
iota_test: Bump test timeout to allow it to run with internal test infra.
#18291 opened Oct 14, 2024 by copybara-service bot Loading…
@pizzud
Pass the compile options for deserialization via PJRT C API
#18287 opened Oct 14, 2024 by jaro-sevcik Loading…
1
[XLA:GPU] Fix includes for reduce_scatter_combiner.cc.
#18285 opened Oct 14, 2024 by copybara-service bot Loading…
@golechwierowicz
[xla:gatherExpander] Extend GatherToLoop transformation to generate correct
#18284 opened Oct 14, 2024 by copybara-service bot Loading…
@bixia1
Integrate LLVM at llvm/llvm-project@1a787b3c8e71
#18276 opened Oct 14, 2024 by copybara-service bot Loading…
@d0k
PR #18062: [ROCm] Fix gemm_rewriter_test for AMD GCN Arch
#18275 opened Oct 14, 2024 by copybara-service bot Loading…
PR #17259: Adding Strictness level to PGLE accuracy checker.
#18273 opened Oct 14, 2024 by copybara-service bot Loading…
[NVIDIA GPU] Add a new option to enable windowed einsum based on total flops of the gemm
#18271 opened Oct 14, 2024 by Tixxx Loading…
Automated Code Change
#18270 opened Oct 14, 2024 by copybara-service bot Loading…
Automated Code Change
#18269 opened Oct 14, 2024 by copybara-service bot Loading…
Automated Code Change
#18268 opened Oct 14, 2024 by copybara-service bot Loading…
Upgrade Abseil to latest LTS branch (lts_2024_07_22_0).
#18267 opened Oct 13, 2024 by copybara-service bot Loading…
ProTip! Filter pull requests by the default branch with base:main.