Commits

Branch selector

 master

User selector

pifon2a

Datepicker

All time

Commit History

Commits on Jan 16, 2022

• [ST] Create directory structure for ST (structured) dialect.

  Show description for 1a46edf
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  1a46edf

  Copy full SHA for 1a46edf

Commits on Jan 12, 2022

• [tfrt:jit] Allow vectorization of 1D reductions outside of linalg.tiled_loop.

  Show description for 445c255
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  445c255

  Copy full SHA for 445c255

• [tfrt:jit] Change tiling of 1D reductions to reduce more vectors per iteration.

  Show description for 82c65f2
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  82c65f2

  Copy full SHA for 82c65f2

• [tfrt:jit] Rewrite 1D vector.multi_reduction differently from N-D.

  Show description for d66f08f
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  d66f08f

  Copy full SHA for d66f08f

Commits on Jan 11, 2022

• [tfrt:jit] Move vector.multi_reduction rewrite to TF CPURT pipeline.

  Show description for 862402f
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  862402f

  Copy full SHA for 862402f

Commits on Jan 10, 2022

• Integrate LLVM at llvm/llvm-project@e745507eda27

  Show description for 4299900
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  4299900

  Copy full SHA for 4299900

Commits on Jan 7, 2022

• [tfrt:jit] Re-enable vectorization of the horizontal reduction.

  Show description for 6fc70d4
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  6fc70d4

  Copy full SHA for 6fc70d4

• Integrate LLVM at llvm/llvm-project@75d65293ca83

  Show description for 3f2215b
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  3f2215b

  Copy full SHA for 3f2215b

• [tfrt:jit] Rename "codegen" to "tiling" passes since they don't do a lot more.

  Show description for 003a90b
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  003a90b

  Copy full SHA for 003a90b

Commits on Jan 6, 2022

• Integrate LLVM at llvm/llvm-project@c5965a411c63

  Show description for 9a58bf0
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  9a58bf0

  Copy full SHA for 9a58bf0

Commits on Jan 5, 2022

• [tfrt:jit] Re-enable vectorization of the horizontal reduction.

  Show description for a0d91a7
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  a0d91a7

  Copy full SHA for a0d91a7

• Integrate LLVM at llvm/llvm-project@b5149f4e66a4

  Show description for dabfc2c
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  dabfc2c

  Copy full SHA for dabfc2c

• [tfrt:jit] Remove sink_unused_outputs pass.

  Show description for 5ca2294
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  5ca2294

  Copy full SHA for 5ca2294

Commits on Jan 4, 2022

• Integrate LLVM at llvm/llvm-project@3064dd8ccffc

  Show description for 49634a9
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  49634a9

  Copy full SHA for 49634a9

• Integrate LLVM at llvm/llvm-project@4c2aba999e54

  Show description for f39ea04
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  f39ea04

  Copy full SHA for f39ea04

• Integrate LLVM at llvm/llvm-project@cb9ccd38c55f

  Show description for 728c67e
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  728c67e

  Copy full SHA for 728c67e

Commits on Jan 3, 2022

• [tfrt:jit] Clean up after rollback.

  Show description for 3eebb23
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  3eebb23

  Copy full SHA for 3eebb23

• Integrate LLVM at llvm/llvm-project@862fffd8231c

  Show description for 3b043f2
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  3b043f2

  Copy full SHA for 3b043f2

• [mlir] Remove redundant reshapes of the `tensor.from_elements` result.

  Show description for 322cba0
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  322cba0

  Copy full SHA for 322cba0

• [KERNEL_GEN] Use `overrideIndexBitwidth(32)` till b/213038246 is fixed.

  Show description for 2c41de7
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  2c41de7

  Copy full SHA for 2c41de7

• Integrate LLVM at llvm/llvm-project@127d95544164

  Show description for f61dc88
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  f61dc88

  Copy full SHA for f61dc88

• Integrate LLVM at llvm/llvm-project@5527139302d9

  Show description for 8b3e2b2
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  8b3e2b2

  Copy full SHA for 8b3e2b2

Commits on Dec 15, 2021

• [KERNEL_GEN] Purge bufferization patterns for RankOp/DimOp.

  Show description for e0910ba
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  e0910ba

  Copy full SHA for e0910ba

Commits on Dec 14, 2021

• [tfrt:jit] Remove `tf_cpurt_` prefix from jit tests and transforms.

  Show description for bacea6d
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  bacea6d

  Copy full SHA for bacea6d

Commits on Dec 13, 2021

• [hlo] Update comments that mention `linalg.tensor_expand/collapse_shape`.

  Show description for a2c5274
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  a2c5274

  Copy full SHA for a2c5274

• [tfrt:jit] Add options to tune linalg transformations.

  Show description for 1a386e3
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  1a386e3

  Copy full SHA for 1a386e3

Commits on Dec 10, 2021

• [tfrt:jit] Split tiling of reductions from the output fusion of FillOp.

  Show description for 79db3c9
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  79db3c9

  Copy full SHA for 79db3c9

Commits on Dec 7, 2021

• [tfrt:jit] Remove padding pass from TF CPURT.

  Show description for d9b47d5
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  d9b47d5

  Copy full SHA for d9b47d5

• [HLO] Fix lhlo-legalize-to-loops pass.

  Show description for e52531a
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  e52531a

  Copy full SHA for e52531a

Commits on Dec 6, 2021

• [tfrt:jit] Add a benchmark for sum(abs(lhs) + exp(rhs)).

  Show description for 6c67d93
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  6c67d93

  Copy full SHA for 6c67d93

• [tfrt:jit] Add tf.Mean benchmark.

  Show description for bb18f63
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  bb18f63

  Copy full SHA for bb18f63

• [tfrt:jit] Update softmax benchmark.

  Show description for 5e207f7
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  5e207f7

  Copy full SHA for 5e207f7

Commits on Dec 2, 2021

• [tfrt:jit] Rewrite reduction benchmarks to use benchmark_mlir_function.

  Show description for 2fadd8c
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  2fadd8c

  Copy full SHA for 2fadd8c

• [tfrt:jit] Detect combiner in linalg.generic reductions during tiling.

  Show description for 4c7969f
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  4c7969f

  Copy full SHA for 4c7969f

Commits on Nov 22, 2021

• [tfrt:jit] Rewrite 1D reductions for vectorization.

  Show description for b1b82cb
  

  pifon2a

  authored and

  tensorflower-gardener

  committed
  b1b82cb

  Copy full SHA for b1b82cb

Pagination

Previous

Next