Google Git
Sign in
opensecura / 3p / openxla / iree / f5660ee72bd8e71910d00cec726375fb4b4a6de8
  1. f5660ee Harden how ConstEval uses llvm-cpu and the runtime libraries. (#17075) by Scott Todd · 12 months ago
  2. d12291f Add Colab notebook showing Hugging Face import via Turbine. (#17093) by Scott Todd · 12 months ago
  3. 36b3ce1 [runtime][cts] Add test to wait multiple times for the same semaphore value (#17125) by Boian Petkantchin · 12 months ago
  4. 46f03af [CodeGen] Add a basic unit test for the TilingConfig class (#17072) by Andrzej Warzyński · 12 months ago
  5. 07d88b1 [CPU][NFC] CPU/KernelDispatch cleanups (#17124) by Benoit Jacob · 12 months ago
  6. 0d947f3 Removing the `iree.compiler.consteval` attr. (#17056) by Ben Vanik · 12 months ago
  7. cbb4257 Update PckgCI cpu testing to include SDXL model testing + benchmark (#17117) by saienduri · 12 months ago
  8. 44ccc22 [runtime][hip][cuda] Fix bug where zombified actions may not get cleaned up (#17107) by Boian Petkantchin · 12 months ago
  9. f4bee0c [runtime] Make the runtime more TSan-friendly (#17051) by Boian Petkantchin · 12 months ago
  10. 0f6bc24 [runtime][hip][cuda] Fix mutex locking when waiting on a semaphore (#17118) by Boian Petkantchin · 12 months ago
  11. afd7cab Add MAINTAINERS.md and RELEASING.md. (#17122) by Stella Laurenzo · 12 months ago
  12. f50de8c Add Python 3.12 to Windows and MacOS builds. by Stella Laurenzo · 12 months ago
  13. fd79fca [Codegen] NFC: Cleanup common transform dialect ops (#17120) by Quinn Dawkins · 12 months ago
  14. d861372 [Codegen] Drop iree.fold_arith_ext_into_contraction for upstream variant (#17119) by Quinn Dawkins · 12 months ago
  15. e8f4948 [DT] Teach encoding about padding. (#17077) by Han-Chung Wang · 1 year ago
  16. 78005ef [runtime][cts] add test where 2 batches wait on different semaphore values (#17091) by Boian Petkantchin · 1 year ago
  17. 0ec9166 Add f32_to_i2 and i2_to_f32 e2e tests. (#17074) by Han-Chung Wang · 1 year ago
  18. 074cbf3 [runtime] Refactor semaphore submission CTS (#17108) by Boian Petkantchin · 1 year ago
  19. cda70e8 Integrate LLVM at llvm/llvm-project@08163cd9d82690e808c28515523b5fd0923d7b38 (#17116) by Stanley Winata · 1 year ago
  20. 36b3891 Skip unused check test compilation in riscv + emscripten jobs. (#17114) by Scott Todd · 1 year ago
  21. 8e86156 [LLVMGPU] Modify layouts to be able to handle dequant operation. (#17113) by Stanley Winata · 1 year ago
  22. e87ff17 [LLVMGPU] allow multiple m and n dims in contraction distribution (#16943) by Quinn Dawkins · 1 year ago
  23. 9f97989 [CodeGen] Fix MLIR types in the function comment. (#17109) by Han-Chung Wang · 1 year ago
  24. 3f51a55 Replace openxla/iree with iree-org/iree across the project. (#17110) by Scott Todd · 1 year ago
  25. 04b9f76 Update instructions for getting write-access in iree-org. (#17112) by Scott Todd · 1 year ago
  26. 6295074 Replace openxla with iree-org in Github runner configs and scripts (#17065) by Jerry Wu · 1 year ago
  27. 5ed2fec [GlobalOptimization] Add a pass to do horizontal fusion of contraction operations with a common operand. (#17059) by Prashant Kumar · 1 year ago
  28. a2476ce [metal] Disable failing semaphore submission test until fixing (#17100) by Lei Zhang · 1 year ago
  29. 0e1e6bf Clarify fusion heuristic (#17098) by MaheshRavishankar · 1 year ago
  30. 125f420 [CodeGen] Add a pattern to fold extract_slice consumer into xfer.write. (#17067) by Han-Chung Wang · 1 year ago
  31. f755b42 [Codegen] Add folding in createBoundedTileSize for partially dynamic wgSize. (#17089) by Stanley Winata · 1 year ago
  32. a0b4853 Fixes to enable out-of-tree plugin builds. (#17095) by MaheshRavishankar · 1 year ago
  33. bd1b106 [CodeGen] Drop encoding for HAL and Flow ops when DT is not supported. (#17081) by Han-Chung Wang · 1 year ago
  34. 886c416 [Winograd] Use TilingInterface for all levels of winograd op tiling (#17061) by Max191 · 1 year ago
  35. 2441959 Add GPU dialect dependencies to C/Python bindings. (#17090) by Scott Todd · 1 year ago
  36. ab4babe [ConstEval] Add flag to adjust tensor size limit for hoisting (#17064) by Max191 · 1 year ago
  37. eec081c [LLVMGPU] Fallback if dynamic dim found on vector distribute. (#17085) by Stanley Winata · 1 year ago
  38. f86f21c [python] Expose MLIR python bindings for gpu and transform (#17088) by Martin Paul Lücke · 1 year ago
  39. 4778f5f Categorize matmul/metvec like generic ops for dispatches (#17084) by Lei Zhang · 1 year ago
  40. f32a87c [Flow] Move elementwise op fusion and bubble up expand shapes patterns into their own pass. (#17068) by MaheshRavishankar · 1 year ago
  41. 3677fbc [runtime] Add semaphore test where 2 batches wait on a former batch amongst 2 (#17080) by Boian Petkantchin · 1 year ago
  42. ff624dd Categorize dispatch name better for linalg.generic cases (#16677) by Lei Zhang · 1 year ago
  43. d284154 Making FlattenFullFillToSplat more conservative. (#17079) by Ben Vanik · 1 year ago
  44. a8731a3 Set top level token permissions (#16744) by Marius Brehler · 1 year ago
  45. cd282de [runtime][hip][cuda] Fix waiting on a semaphore on the host (#17073) by Boian Petkantchin · 1 year ago
  46. fdfe344 [LinalgExt] Moving encoding utils to EncodingAttr builtin or LinalgExt/IR (#17053) by Han-Chung Wang · 1 year ago
  47. c83f9ba NFC: Add VectorLayoutInterface method for getting the layout rank (#17071) by Quinn Dawkins · 1 year ago
  48. 5adae9e [Codegen] Fix `TilingConfig::getTilingLevelForVectorDimPosition()` for <= 2 tiling levels (#17050) by Benjamin Maxwell · 1 year ago
  49. ace6397 Add boolean option to do RELU in mlp plugin (#17058) by Nirvedh Meshram · 1 year ago
  50. 872f0b6 Fix crash in transform dialect script when using attention script with ToT IREE (#17066) by MaheshRavishankar · 1 year ago
  51. 56541e4 [NFC] Switching to not use using-directives in UtilExternalModels.cpp (#17063) by Han-Chung Wang · 1 year ago
  52. 5e75105 [Flow] Implement ValueBoundsOpInterface for flow.dispatch.tensor.load op (#17062) by Han-Chung Wang · 1 year ago
  53. 06f41ce [Preprocessing] Add pad to MMA intrinsic size pass (#17057) by Jakub Kuderski · 1 year ago
  54. 6d4a99c Use `iree-turbine` in PyTorch docs and samples. (#17036) by Scott Todd · 1 year ago
  55. 915b42e Add DataLayoutPropagation pass to bubble up/push down pack and unpack (#16731) by Jerry Wu · 1 year ago
  56. 503edab [GlobalOpt][DT] Simplify logics in SetEncoding pass. (#17040) by Han-Chung Wang · 1 year ago
  57. 0bd3c1d [stablehlo] Update stablehlo to 341e063f0924fc1350538dc53a92c21ec5e022a3 (#17026) by Balaji V. Iyer · 1 year ago
  58. e36844f Add ability to call the same custom dispatch multiple times when using pdl patterns. (#16967) by Nirvedh Meshram · 1 year ago
  59. 1d00b50 Update Discord invite link. (#17052) by Scott Todd · 1 year ago
  60. 22faa15 Update README.md with other discord link. by Stella Laurenzo · 1 year ago
  61. 72aeee2 Update README.md with new Discord link. by Stella Laurenzo · 1 year ago
  62. 2039d56 Remove the fixed point iteration in the global opt pipeline. (#17049) by Ben Vanik · 1 year ago
  63. 529826f Add missing line continuation slash to recently updated page. (#17048) by Scott Todd · 1 year ago
  64. 459fab6 [runtime][hip][cuda] Fix waiting on wait semaphores before executing actions (#17025) by Boian Petkantchin · 1 year ago
  65. 3d626b5 [Preprocessing] NFC: Finish migrating passes to use new tablegen (#17047) by Quinn Dawkins · 1 year ago
  66. 39091a7 [Flow] Switch to new pass generation tablegen definitions (#17046) by Quinn Dawkins · 1 year ago
  67. 1c49d6a [runtime][hip][cuda] Add tracing in graph execution mode (#16894) by Boian Petkantchin · 1 year ago
  68. 080657f Fix failing transform dialect CUDA tests. (#17042) by MaheshRavishankar · 1 year ago
  69. bdbf42a Clone `tensor.empty` operations on dispatches with only `linalg.generic` ops. (#17043) by MaheshRavishankar · 1 year ago
  70. c2abb93 Disable TD tests on CUDA backends due to failure. (#17041) by MaheshRavishankar · 1 year ago
  71. 55fafcf Forking dynamic behavior from flow.tensor.constant. (#17034) by Ben Vanik · 1 year ago
  72. 954cb36 Move Codegen pass pipelines to nest on `FunctionOpInterface`. (#16665) by MaheshRavishankar · 1 year ago
  73. 699d244 Integrate llvm 20240412 (#17031) by Vivian · 1 year ago
  74. f616123 [python] Expose python bindings for amdgpu in iree.compiler.dialects (#17028) by Martin Paul Lücke · 1 year ago
  75. 40f2533 [HIP] Add inline execution mode (#16951) by Nithin Meganathan · 1 year ago
  76. 390f865 Integrate llvm 20240411 at 56954a53e58282d7584e31ec14a2b1052cd861e8 (#17027) by Vivian · 1 year ago
  77. 76e9cfe Allowing flow.tensor.constant to be used for constants. (#17024) by Ben Vanik · 1 year ago
  78. fbd31b0 [Vulkan] Fix coop matrix property initialization (#17023) by Jakub Kuderski · 1 year ago
  79. 94971b4 [LLVMGPU] Fit mma schedules inside shared memory limits (#16927) by Kunwar Grover · 1 year ago
  80. 4437c43 [CPU RISCV] Avoid using pre-configured tile sizes as input vector sizes (#17018) by Bruce Lai · 1 year ago
  81. 26f77de [CPU] Fix FusionOfTensorOps nullptr (#17015) by Diego Caballero · 1 year ago
  82. 7dca44b [GPU] Overhaul reduce shared memory bank conflicts pass for gfx9 (#17010) by Jakub Kuderski · 1 year ago
  83. 0a92a71 Integrate llvm at 9760872b537ba8e6eee2e68eb81b7d26af5b40e4 (#17011) by Vivian · 1 year ago
  84. 2566f15 Re-enable ROCm ONNX tests, running one test at a time. (#17014) by Scott Todd · 1 year ago
  85. 67e234c Update "bindings" reference page to reflect current support status. (#17005) by Scott Todd · 1 year ago
  86. 2780fd5 Fix typo in benchmark tracing warning message. by Ben Vanik · 1 year ago
  87. 336ba12 Fill in details on parameter formats (IRPA, GGUF, safetensors). (#17006) by Scott Todd · 1 year ago
  88. e006a05 Document recently added utility compiler passes. (#16983) by Scott Todd · 1 year ago
  89. b4273a4 Increase pip install timeout for pkgci venv setup. (#17008) by Scott Todd · 1 year ago
  90. 92f10a7 [CodeGen] Add fpowi pattern to PolynomialApproximationPass (#17003) by Daniel Garvey · 1 year ago
  91. 8a94d5c Refresh "profiling with Tracy" developer docs. (#16939) by Scott Todd · 1 year ago
  92. 8ee0ada [GlobalOpt][DT] Add a flag to disable early materialization. (#16997) by Han-Chung Wang · 1 year ago
  93. d42e457 Add e2e tests for FA2 (#16953) by erman-gurses · 1 year ago
  94. ab949ef [Codegen] Add support for vectorizing tensor.unpack ops with masking. (#16664) by Han-Chung Wang · 1 year ago
  95. 5a95fd4 Avoid distributing loops that are statically known to be unit trip count (#16985) by MaheshRavishankar · 1 year ago
  96. a144ff6 Fix a “Library not loaded” issue on macos (#16987) by Atomoper · 1 year ago
  97. 190d959 Allow users to specify riscv cpu and get hardware features (#16902) by Alex Chiang · 1 year ago
  98. bb7e536 Disable out of tree ROCm tests again. (#16994) by Scott Todd · 1 year ago
  99. dcc8e19 Adds a flag to enable/disable vector contract custom kernels in `LLVMCPUMmt4dVectorLoweringPass` (#16867) by Kojo Acquah · 1 year ago
  100. 39bf204 Integrate LLVM at 9708d0900311503aa4685d6810d8caf0412e15d7 (#16988) by Benoit Jacob · 1 year, 1 month ago