Pull requests: tensorflow/tensorflow
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
Lower index to 64 bits if 64 bit indices are used in HLO.
#66832
opened May 2, 2024 by
copybara-service
bot
•
Draft
Add a test case to document how HloCSE works for multi-output fusions.
#66831
opened May 2, 2024 by
copybara-service
bot
•
Draft
[XLA:GPU] Add an optional compiler pass that verifies Triton fusions.
#66830
opened May 2, 2024 by
copybara-service
bot
•
Draft
[XLA:GPU] Cast bf16 matmuls to fp32 on pre-Ampere GPUs
#66825
opened May 2, 2024 by
copybara-service
bot
•
Draft
[XLA:GPU][MLIR-based emitters] Remove unused dimensions in xla_gpu.indexing_op.
#66824
opened May 2, 2024 by
copybara-service
bot
•
Draft
[StablehloShapeRefinement] Skip constant folding of convert operations with dynamic shapes.
#66822
opened May 2, 2024 by
copybara-service
bot
•
Draft
Add a check that we don't have duplicate tuple elements.
#66820
opened May 2, 2024 by
copybara-service
bot
•
Draft
Reverts 08fdea60bee843ec228eacef8bb37fc62b87108f
#66817
opened May 2, 2024 by
copybara-service
bot
•
Draft
Enable saving quantization report to file.
#66816
opened May 2, 2024 by
copybara-service
bot
•
Draft
Add TFExecutorGraphPruningPass to TF dialect to executor export pipeline
#66815
opened May 2, 2024 by
copybara-service
bot
•
Draft
[PJRT] Remove the deprecated PjRtClient::LookupDevices() overload that takes a raw int.
#66811
opened May 2, 2024 by
copybara-service
bot
•
Draft
[XLA:GPU] Enable strength-reduction for dots that are not supported by GemmFusion
#66809
opened May 2, 2024 by
copybara-service
bot
•
Draft
[xla:gpu] ExecutionCounters for Send/Recv thunks should be per StreamExecutor
#66807
opened May 1, 2024 by
copybara-service
bot
•
Draft
[xla:cpu] NFC: Remove deprecated XLA:CPU mlir based codegen part #4
#66805
opened May 1, 2024 by
copybara-service
bot
•
Draft
Previous Next
ProTip!
What’s not been updated in a month: updated:<2024-04-02.