-
Notifications
You must be signed in to change notification settings - Fork 476
Pull requests: openxla/xla
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
[xla:cpu:xnn] Do not use XNNPACK for dots that require tiling by K
#21591
opened Jan 18, 2025 by
copybara-service
bot
Loading…
[xla:cpu:xnn] Do not return XNN runner back to the pool until execution is complete
#21590
opened Jan 18, 2025 by
copybara-service
bot
Loading…
[xla:emitters] flatten_tensors: CPU support
#21589
opened Jan 18, 2025 by
copybara-service
bot
Loading…
[xla:emitters] lower_tensors: initial CPU support
#21588
opened Jan 18, 2025 by
copybara-service
bot
Loading…
[pjrt] Use the
PjRtMemorySpace*
version of BufferFromHostLiteral
#21587
opened Jan 18, 2025 by
copybara-service
bot
Loading…
[xla:cpu:xnn] Measure execution time of parallel task to decide the optimal number of workers
#21580
opened Jan 18, 2025 by
copybara-service
bot
Loading…
Add cache entries for reshape ops in SPMD.
#21579
opened Jan 18, 2025 by
copybara-service
bot
Loading…
TfrtCpuClient
should respect run_backend_only
option.
#21578
opened Jan 18, 2025 by
copybara-service
bot
Loading…
[JAX] Include the memory kind when converting JAX/IFRT sharding types
#21577
opened Jan 18, 2025 by
copybara-service
bot
Loading…
[XLA:Python] Remove unused GOOGLE_CUDA/TENSORFLOW_USE_ROCM macro guarding a header inclusion.
#21576
opened Jan 18, 2025 by
copybara-service
bot
Loading…
Previous Next
ProTip!
Adding no:label will show everything without a label.