issues
search
openxla
/
xla
A machine learning compiler for GPUs, CPUs, and ML accelerators
Apache License 2.0
2.39k
stars
358
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
[xla:cpu] Add FFI custom call thunk runtime support to PJRT CPU client.
#14288
copybara-service[bot]
closed
4 days ago
0
[XLA:CPU] Port concatenate instruction to Thunks
#14287
copybara-service[bot]
opened
4 days ago
1
[IFRT] Include input_specs in RemapPlan::DebugString()
#14286
copybara-service[bot]
closed
4 days ago
0
[XLA:FFI] Catch exceptions in user FFI calls.
#14285
copybara-service[bot]
closed
1 day ago
0
Clean up tile size computation for row reductions.
#14284
copybara-service[bot]
closed
4 days ago
0
Break circular dependency between gpu_stream.cc and gpu_event.cc.
#14283
copybara-service[bot]
closed
4 days ago
0
Unify indexing map computation for reduction emitters.
#14282
copybara-service[bot]
closed
4 days ago
0
Move client creation to a separate file (NFC).
#14281
copybara-service[bot]
closed
2 days ago
0
Standardize patch format to work with internal tooling.
#14280
copybara-service[bot]
opened
4 days ago
0
Clean up some duplication and dead code in reduction emitter.
#14279
copybara-service[bot]
closed
4 days ago
0
Reverts 8e7c93698dd15a3c6c347c301de1875dff612515
#14278
copybara-service[bot]
closed
4 days ago
0
Remove unused GetBinaryDir function.
#14277
copybara-service[bot]
closed
4 days ago
0
[XLA] Prohibit defining `Literal::AbslHashValue` as it cannot be made consistent.
#14276
copybara-service[bot]
closed
4 days ago
0
[XLA:GPU] Move `triton_gpu.sparse_dot` to LLVM pattern from Triton patch (`convert-triton-gpu-to-llvm` pass) to OpenXLA (`sparse-convert-layout-op` pass).
#14275
copybara-service[bot]
closed
4 days ago
0
[XLA:GPU] Disable cublasLT tracing for cuda version < 12030
#14274
shawnwang18
closed
4 days ago
1
Reverts bf95473bdb484f742de0df899329f15b3e2249ed
#14273
copybara-service[bot]
closed
4 days ago
0
Explicitly disallow duplicated devices during array construction
#14272
copybara-service[bot]
closed
4 days ago
0
Automated Code Change
#14271
copybara-service[bot]
opened
5 days ago
0
Don't refactor dependencies into a bzl file.
#14270
copybara-service[bot]
closed
4 days ago
0
PR #14166: Add kPower case to algsimp IsNonNegative
#14269
copybara-service[bot]
opened
5 days ago
0
[PJRT:GPU] Implement copying buffers to pinned host memory space
#14268
jaro-sevcik
closed
4 days ago
1
Make GPU PJRT client error out early when not compiled with GPU support
#14267
copybara-service[bot]
opened
5 days ago
0
[XLA:GPU] Add command buffer custom call targets recording for legacy custom call registry API
#14266
shawnwang18
closed
20 hours ago
6
[IFRT] Modify ifrt-verify-donation to reject instances when an arg is both donated and not donated.
#14265
copybara-service[bot]
closed
4 days ago
0
[XLA:GPU] Add debug info for command buffer trace cache
#14264
shawnwang18
opened
5 days ago
1
[xla:ffi] Add an API to update CallFrame in place
#14263
copybara-service[bot]
closed
5 days ago
0
[xla:ffi] NFC: Implement Update as Clone and UpdateInPlace
#14262
copybara-service[bot]
closed
5 days ago
0
[xla:ffi] NFC: Use absl::InlinedVector to store dimensions
#14261
copybara-service[bot]
closed
5 days ago
0
[xla:ffi] Add an API to update CallFrame with new run time values (buffer pointers)
#14260
copybara-service[bot]
closed
5 days ago
0
Downgrade TSL's zlib to fix breakage
#14259
copybara-service[bot]
closed
5 days ago
0
[XLA:CollectivePipeliner] Sort formatting_ops (wrt original order) so that the mapper can traverse them in a way that operands are already mapped.
#14258
copybara-service[bot]
closed
5 days ago
0
[tsl::mutex] Implement `assert_held()` and `assert_held_shared()` methods.
#14257
copybara-service[bot]
closed
5 days ago
0
Removed redundant *Internal methods from CpuCallback
#14256
copybara-service[bot]
closed
4 days ago
0
Fix build issue in xla/hlo/utils/hlo_sharding_util_test.cc
#14255
apivovarov
opened
5 days ago
4
Fix symbolic_tile_analysis_test.cc build
#14254
apivovarov
closed
5 days ago
2
[XLA] Clean up and modernize tuple_simplifier and corresponding test suite.
#14253
copybara-service[bot]
closed
4 days ago
0
Add TPU v5 to `c_api_decl.h` and `tpu_topology.cc`
#14252
copybara-service[bot]
closed
5 days ago
0
[XLA:Python] Tiny optimization in traceback hashing.
#14251
copybara-service[bot]
closed
19 hours ago
0
Change Shardonnay to Shardy
#14250
copybara-service[bot]
opened
5 days ago
0
PR #14158: Add kExp op to algsimp IsNonNegative
#14249
copybara-service[bot]
closed
5 days ago
0
Use `if_google` instead of buildozer transformations in `auto_sharding`
#14248
copybara-service[bot]
closed
5 days ago
0
[XLA] Make sure cosh does not return values < 1
#14247
copybara-service[bot]
closed
4 days ago
0
Update version of nsync used by TensorFlow to version 1.29.2
#14246
copybara-service[bot]
closed
1 day ago
0
[xla:ffi] Optimize CallFrame construction
#14245
copybara-service[bot]
closed
5 days ago
0
[XLA:GPU] Refactor tree_reduction_rewriter into 3 parts.
#14244
copybara-service[bot]
opened
5 days ago
0
[mutex] Add `assert_held` and `assert_shared_held` methods.
#14243
copybara-service[bot]
opened
5 days ago
0
[XLA:FFI] Reduce the cost of FFI CallFrame creation and destruction.
#14242
copybara-service[bot]
opened
5 days ago
0
[xla:cpu] Move convolution implementation file to the 'runtime' directory
#14241
copybara-service[bot]
closed
4 days ago
1
[XLA:Python] Remove code to support Python 3.9.
#14240
copybara-service[bot]
closed
5 days ago
0
[xla:cpu] Add convolution shape verification
#14239
copybara-service[bot]
closed
1 day ago
1
Previous
Next