Commits

Branch selector

 2f77580

User selector

All users

Datepicker

All time

Commit History

Commits on Sep 21, 2022

• Update VM image setup scripting (iree-org#10489)

  Show description for 2f77580
  GMNGeoffrey

  authored
  2f77580

  Copy full SHA for 2f77580

• Cleanup markers at the end of TileAndDistribute pass (iree-org#10492)

  Show description for ac81262
  ThomasRaoux

  authored
  ac81262

  Copy full SHA for ac81262

• [NFC] Replace legacy linalg.pad_tensor ops with tensor.pad ops (iree-org#10486)

  Show description for 998568e
  hanhanW

  authored
  998568e

  Copy full SHA for 998568e

• Move benchmark suites into python dir (iree-org#10415)

  

  Jerry Wu

  authored
  71562cc

  Copy full SHA for 71562cc

• Fix iree_bytecode_module RISCV crosscompile flags (iree-org#10471)

  Show description for d27d6b4
  hcindyl

  authored
  d27d6b4

  Copy full SHA for d27d6b4

• Fix timeout property in iree_native_test (iree-org#10470)

  hcindyl

  authored
  f2d4edb

  Copy full SHA for f2d4edb

Commits on Sep 20, 2022

• Add an option to control on applying tensor.pad -> insert_slice pattern. (iree-org#10435)

  Show description for caa7842
  hanhanW

  authored
  caa7842

  Copy full SHA for caa7842

• Update default vector level tiling sizes for ARM and RISC-V depthwise conv. (iree-org#10454)

  Show description for a5792bf
  hanhanW

  authored
  a5792bf

  Copy full SHA for a5792bf

• rename iree_ukernel_size_t to iree_ukernel_ssize_t (iree-org#10484)

  Show description for 938c32b
  bjacob

  authored
  938c32b

  Copy full SHA for 938c32b

• Remove workload per workgroup (iree-org#10464)

  Show description for 2d1143d
  okkwon

  authored
  2d1143d

  Copy full SHA for 2d1143d

• mmt4d i8mm kernel register fix (iree-org#10476)

  Show description for 9b97d80
  bjacob

  authored
  9b97d80

  Copy full SHA for 9b97d80

• [LLVMCPU] Do not return the dim size if it is greater than maxSize. (iree-org#10437)

  Show description for 8ad7a71
  hanhanW

  authored
  8ad7a71

  Copy full SHA for 8ad7a71

• Unpeeled Epilogue for WMMA-based GEMMs (iree-org#10451)

  Show description for c029c24
  

  Manish Gupta

  authored
  c029c24

  Copy full SHA for c029c24

• [vulkan] Create zero binding for dummy sets in BuiltinExecutables (iree-org#10448)

  Show description for e58f407
  antiagainst

  authored
  e58f407

  Copy full SHA for e58f407

• [spirv] Add support for user specified configuration (iree-org#10430)

  Show description for 5a31f58
  harsh-nod

  authored
  5a31f58

  Copy full SHA for 5a31f58

• Rationalize architecture checks in ukernels (iree-org#10475)

  Show description for 165949e
  bjacob

  authored
  165949e

  Copy full SHA for 165949e

• Add elementwise fusion on multiuse ops (iree-org#10446)

  Show description for 4cbcaaf
  

  Jerry Wu

  and

  Mahesh Ravishankar

  authored
  4cbcaaf

  Copy full SHA for 4cbcaaf

Commits on Sep 19, 2022

• Commit to check funtionality and performance for wmma with unpeeled epilogue

  

  Manish Gupta

  committed
  bff857d

  Copy full SHA for bff857d

• Don't promote matmul for gemv kind of case (iree-org#10477)

  Show description for 46dc79c
  ThomasRaoux

  authored
  46dc79c

  Copy full SHA for 46dc79c

• Generalize Transpose to Support 3D Shapes with Batch Dimensions (iree-org#10393)

  Show description for 32e8a89
  KoolJBlack

  authored
  32e8a89

  Copy full SHA for 32e8a89

• Add new models to compare against TFLite (iree-org#10421)

  mariecwhite

  authored
  5c8c00c

  Copy full SHA for 5c8c00c

• detect cpu features on Apple/ARM64 (iree-org#10469)

  Show description for 4a00384
  bjacob

  authored
  4a00384

  Copy full SHA for 4a00384

• Teach ResolveBufferDescriptors how to handle memref.get_global. (iree-org#10459)

  Show description for 9f8abc4
  

  Stella Laurenzo

  authored
  9f8abc4

  Copy full SHA for 9f8abc4

• Fix GPU shared memory vector copy possibility analysis (iree-org#10468)

  Show description for 4e3db5e
  antiagainst

  authored
  4e3db5e

  Copy full SHA for 4e3db5e

• Revert "Only enable CUDA benchmarks when IREE_TARGET_BACKEND_CUDA is enabled" (iree-org#10473)

  Show description for 7ba7a81
  

  Jerry Wu

  authored
  7ba7a81

  Copy full SHA for 7ba7a81

• [spirv] Update default AMD convolution and matmul configuration (iree-org#10467)

  Show description for cb03865
  antiagainst

  authored
  cb03865

  Copy full SHA for cb03865

Commits on Sep 18, 2022

• Only enable CUDA benchmarks when IREE_TARGET_BACKEND_CUDA is enabled (iree-org#10457)

  

  Jerry Wu

  authored
  5e61265

  Copy full SHA for 5e61265

• Reactivate reduction test and improve strategy. (iree-org#10445)

  Show description for 6e5728f
  nicolasvasilache

  authored
  6e5728f

  Copy full SHA for 6e5728f

• mmt4d builtin ukernels: baseline ARM64 kernels for f32 and i8 (iree-org#10440)

  Show description for 77e7cac
  bjacob

  authored
  77e7cac

  Copy full SHA for 77e7cac

Commits on Sep 17, 2022

• Report registered backends on failure (iree-org#10452)

  Show description for 6c98ebb
  jpienaar

  authored
  6c98ebb

  Copy full SHA for 6c98ebb

• Fix RISC-V 64 Linux test script (iree-org#10461)

  Show description for b64bd64
  hcindyl

  authored
  b64bd64

  Copy full SHA for b64bd64

• Integrate llvm-project at llvm/llvm-project@42a90e6017b0 (iree-org#10453)

  Show description for a930860
  

  MaheshRavishankar

  and

  Stella Laurenzo

  authored
  a930860

  Copy full SHA for a930860

Commits on Sep 16, 2022

• Support GEMM Pipelining *without* Epilogue Peeling (iree-org#10388)

  

  Manish Gupta

  authored
  8d1e6ab

  Copy full SHA for 8d1e6ab

• fix gcc build (iree-org#10450)

  ThomasRaoux

  authored
  4c7e06a

  Copy full SHA for 4c7e06a

• Disable ARM64 ukernels on APPLE for now (iree-org#10449)

  bjacob

  authored
  250725b

  Copy full SHA for 250725b

Pagination

Previous

Next