Default Branch

master b5d8801b1e · [XLA][HLO Value Tracking] Maintain OriginalValue consistency in MemorySpaceAssignment · Updated 2026-09-26 06:56:09 +08:00

ZIP  TAR.GZ

Branches

exported_pr_989027196 5fe8c21e7f · Automated Code Change · Updated 2026-09-27 13:06:10 +08:00	0 1		ZIP  TAR.GZ
exported_pr_989027201 fabdb6aa5c · Automated Code Change · Updated 2026-09-27 13:05:16 +08:00	0 1		ZIP  TAR.GZ
exported_pr_989003734 f59cbd73c8 · Automated Code Change · Updated 2026-09-27 12:57:05 +08:00	0 1		ZIP  TAR.GZ
exported_pr_988988484 ae4c06b291 · Automated Code Change · Updated 2026-09-27 11:22:12 +08:00	0 1		ZIP  TAR.GZ
exported_pr_988986702 d499b9266c · Automated Code Change · Updated 2026-09-27 11:17:55 +08:00	0 1		ZIP  TAR.GZ
exported_pr_989002184 fdc798334b · Automated Code Change · Updated 2026-09-27 11:17:30 +08:00	0 1		ZIP  TAR.GZ
exported_pr_988989952 930156ac75 · Automated Code Change · Updated 2026-09-27 10:34:43 +08:00	0 1		ZIP  TAR.GZ
exported_pr_988893924 743135d3f3 · Remove redundant requires-gpu-nvidia tags from tf_cuda_cc_test targets · Updated 2026-09-27 09:45:34 +08:00	0 1		ZIP  TAR.GZ
exported_pr_988941986 70fc8cebc6 · Remove the fixed thread cap on DefaultPool() · Updated 2026-09-27 08:07:23 +08:00	0 1		ZIP  TAR.GZ
exported_pr_988368914 c8fb6d3073 · Validate contracting and batch dimensions in XLA Einsum and BatchDot to reject invalid operations with mismatched inner dimensions. · Updated 2026-09-27 06:22:12 +08:00	0 1		ZIP  TAR.GZ
exported_pr_988627339 3aa665471d · Share CUB sort FFI bindings and helpers between CUDA and ROCm. · Updated 2026-09-27 03:34:09 +08:00	0 1		ZIP  TAR.GZ
exported_pr_988629862 00db1641e8 · Use absl::c_* container algorithms in xla/service and xla/hlo/ir. · Updated 2026-09-27 03:32:24 +08:00	0 1		ZIP  TAR.GZ
exported_pr_988606993 963b188ad9 · Fix missing derived-instruction setup on GTEs in AllToAllDecomposer and use HloInstruction::AddInstruction for derived HLOs. · Updated 2026-09-27 03:29:34 +08:00	0 1		ZIP  TAR.GZ
exported_pr_988591232 dcb1123978 · Build TFLite kernel tests without TensorFlow in CMake · Updated 2026-09-27 03:06:19 +08:00	0 1		ZIP  TAR.GZ
exported_pr_988883011 4650630acd · Infer zero-sized VALID pooling outputs · Updated 2026-09-27 01:44:19 +08:00	0 1		ZIP  TAR.GZ
exported_pr_978935069 8c2931b13b · Point gpu-any / nvgpu_any and tf_gpu_tests_tags() at requires-gpu-sm90 (and · Updated 2026-09-27 01:36:04 +08:00	0 1		ZIP  TAR.GZ
exported_pr_983271872 d51a04ac79 · Cast bfloat16 logits to float32 on GPU in stateless_random_ops_test · Updated 2026-09-27 01:29:40 +08:00	0 1		ZIP  TAR.GZ
exported_pr_980526939 d3f08a1f86 · [thunky] Add jax.experimental.thunky for XLA:GPU ThunkExecutor programming. · Updated 2026-09-27 01:09:29 +08:00	0 1		ZIP  TAR.GZ
exported_pr_984912184 3af492906b · Fix Python 3 compatibility in huge_slice_op_test · Updated 2026-09-27 00:38:46 +08:00	0 1		ZIP  TAR.GZ
exported_pr_984912142 c6347554a8 · Cast bfloat16 logits to float32 on GPU in stateless_random_ops_test · Updated 2026-09-27 00:36:49 +08:00	0 1		ZIP  TAR.GZ