Commit Graph

Select branches

Hide Pull Requests

master

Mono Color

• f1a503a610 add extern include in build of gentable master colin 2020-10-14 12:01:16 +0800
• 70a7733278 refine tdfiles in interface colin 2020-10-14 11:53:46 +0800
• 4aa456585e refine onnx build colin 2020-10-14 11:18:48 +0800
• f74023467a remove unuse include colin 2020-10-13 20:04:26 +0800
• e3ad2a5982 refine llvm dep colin 2020-10-13 20:04:09 +0800
• a0f2513706 add bazel support colin 2020-10-13 17:19:49 +0800
• 1b2fae016f modify build in dialect colin 2020-10-12 15:14:29 +0800
• 8af23c97e9 add build in dialect onnx colin 2020-10-10 21:05:48 +0800
• c440bd4dba add bazel support 1/10 colin 2020-09-29 17:32:28 +0800
• 6bd9471262
  Lower ReduceMean op to Krnl dialect (#318) Tung D. Le 2020-10-11 11:09:42 +0900
• 81c774ba5b
  Cleanup rtmemref api (#238) gongsu832 2020-10-10 10:32:09 -0400
• 533d47acf1
  Normalize memrefs with affine_map in krnl.getref (#327) Haruki Imai 2020-10-10 05:59:52 +0900
• 7bfb5c93c1
  Enable bundling of dynamic memory pools on a block basis. (#330) Gheorghe-Teodor Bercea 2020-10-05 13:55:46 -0400
• 931127c7e9
  Lower tile to Krnl (#308) chentong319 2020-10-05 00:50:59 -0400
• cb3d1e4f64
  Import graph output type from protobuf (#333) daquexian 2020-10-03 23:21:15 +0800
• 0db735f48d
  Support per-block bundling for static memory pools. (#325) Gheorghe-Teodor Bercea 2020-10-01 14:15:03 -0400
• 75930ffbcf
  Fix substitution order in parse_type_str (#316) G. Ramalingam 2020-09-30 02:36:06 -0700
• 46306a8a26
  Improve shape inference of flatten Op (#314) chentong319 2020-09-29 12:59:01 -0400
• aa2eed411f
  Remove fork when redirecting stderr in onnx-mlir (#310) NathanielMcVicar 2020-09-25 15:39:46 -0700
• 278f10c3a4
  Normalize memrefs with affine_map in krnl.memcpy (#322) Haruki Imai 2020-09-26 03:36:15 +0900
• 71b8ca28aa
  Fix Debug mode for ONNX-MLIR on Windows (#311) Chun-Wei Chen 2020-09-25 06:58:03 -0700
• 6500f27f80
  Use ubuntu focal as ppc64le base image. (#312) Tian Jin 2020-09-25 16:13:22 +0700
• f0c5b99229
  Gather: fix for negative indices (#313) Alexandre Eichenberger 2020-09-24 14:02:49 -0400
• f16983a7c4
  Update onnx_location.mlir (#319) Tian Jin 2020-09-24 21:17:29 +0700
• 17383768df
  Initial Location info support (#302) Kevin O'Brien 2020-09-23 18:58:27 -0400
• 4bbe12ff50
  Improve support for krnl.dim (#317) Gheorghe-Teodor Bercea 2020-09-23 14:36:16 -0400
• 3491b90b1e
  Support LLVM as of 7dcd0042 (#309) NathanielMcVicar 2020-09-22 09:42:50 -0700
• 4cc16aceb7
  [MLIR] Add SizeOp conversion from ONNX dialect to Krnl dialect (#295) Prashant Kumar 2020-09-21 15:25:21 +0530
• 3520dbd6e1
  temporarily allow ppc64le build to fail (#307) Kevin O'Brien 2020-09-20 09:12:49 -0400
• 66074da3ac
  Lower ONNXConstantOfShapeOp to Krnl dialect (#296) Tung D. Le 2020-09-20 01:47:39 +0900
• 3a5aa7ee31
  Gather ONNX to Kernel Lowering (#294) Alexandre Eichenberger 2020-09-11 15:36:23 -0400
• fa04c32a0c
  ShapeInference for SizeOp (#299) chentong319 2020-09-11 13:47:11 -0400
• 1fcf97ef8d
  initial code for handling custom ops (#288) Kevin O'Brien 2020-09-11 10:49:15 -0400
• ac67900baf
  Rewrite shape and size OP (#285) chentong319 2020-09-10 13:46:00 -0500
• 5e11429d77
  Dropout elimination & Conv Bugfix (#297) Tian Jin 2020-09-10 14:47:30 +0800
• 03dae57189
  Using onnx-mlir through incremental stages (#257) Kevin Wu 2020-09-09 21:29:55 -0500
• dbc41d2330
  Update llvm commit ID to 1d01fc1 (#292) Tian Jin 2020-09-09 23:12:01 +0800
• c738e0fb07
  Fix wrong type for llvm::loadop (#293) Tung D. Le 2020-09-04 23:29:00 +0900
• 9f69b2f317
  Emit the dynamic memory pool (#290) Gheorghe-Teodor Bercea 2020-09-03 10:31:06 -0400
• 930e20f682
  Syntax highlighting for mlir code in README (#276) Tung D. Le 2020-09-02 20:24:41 +0900
• c1262c184e
  NFC: Attribute cleanup (remove references of attributes) (#286) Alexandre Eichenberger 2020-08-31 14:28:16 -0400
• 8bfde7de4b
  Transpose optimization (#280) Alexandre Eichenberger 2020-08-31 10:40:17 -0400
• 8e3748d4c7
  Update available python3 version on AzurePipelines (#279) Chun-Wei Chen 2020-08-28 07:39:20 -0700
• 24d0a2ac71
  Add shape inference and names (#266) Aman LaChapelle 2020-08-27 12:46:27 -0700
• 11a5029c10
  change tag to x86 (#273) Kevin O'Brien 2020-08-26 14:14:59 -0400
• f278f08120
  Introduce krnl.shape operation and its lowering (#267) Gheorghe-Teodor Bercea 2020-08-19 12:57:40 -0400
• 13e8070708
  Make krnl-to-llvm patterns reusable (#272) Tung D. Le 2020-08-20 00:58:04 +0900
• fe6dc5bd81
  Move gtest to submodule (#262) NathanielMcVicar 2020-08-18 05:41:19 -0700
• 4ed95319db
  Fix build config validation on Travis CI (#256) Masahiro H 2020-08-18 18:55:14 +0900
• 7c1e67898d
  Fuse convolution and batch normalization (#253) Tung D. Le 2020-08-18 17:41:40 +0900
• 38bd77e51a
  Build llvm nightly prereq images & test onnx-mlir against it regularly. (#268) Tian Jin 2020-08-18 08:59:50 +0800
• 09205f40f0
  Update .travis.yml (#264) Tian Jin 2020-08-16 16:45:33 +0800
• 00299910f3
  OneHotEncoder Shape Inference (#265) Anh Leu 2020-08-14 15:13:31 -0500
• d3dcee7366
  Add krnl.dim operation and lowering pass (#261) Gheorghe-Teodor Bercea 2020-08-14 12:59:33 -0400
• 1b42d0b4eb
  Update LLVM commit ID to the version that corresponds to MLIR News, 13th edition (8/7/2020) (#248) Tung D. Le 2020-08-14 13:52:48 +0900
• c7b5eecc7f
  Added builder case for int8 tensors in the dialect (#263) Lsz 2020-08-13 06:16:27 -0700
• 2f41c2bf5b
  Assorted fixes for memory pool related passes (#244) Gheorghe-Teodor Bercea 2020-08-11 17:34:59 -0400
• 2ee725d939
  Add CastOp lowering (#259) Anh Leu 2020-08-11 15:07:13 -0500
• e1386b0689
  Add shape inference for Ops used by BERT (#249) albertomagni-ms 2020-08-07 06:08:00 +0100
• 20686d8f0f
  Catch build failure on Windows-CI (Azure Pipelines) (#241) Chun-Wei Chen 2020-08-06 01:53:59 -0700
• dbe0d734b5
  Call LLVM opt to optimize bitcode (#245) Tung D. Le 2020-08-03 12:46:21 +0900
• 58ee62fb49
  [NFC] Rename passes for stylistic consistency. (#232) Tian Jin 2020-07-31 21:37:35 +0800
• b4228fd288
  Seq type (#199) chentong319 2020-07-31 08:05:59 -0400
• 8e8f894574
  Update CMakeLists.txt (#240) Tian Jin 2020-07-31 16:36:03 +0800
• 029fb5eb67
  Add support for emitting individual memory pools with dynamic sizes (#211) Gheorghe-Teodor Bercea 2020-07-30 12:24:07 -0400
• 4db3edc025
  include stdlib.h on macOS (#236) Masahiro H 2020-07-30 17:48:31 +0900
• 7a2bb1f2a3
  Fix mainutils win32 (#239) gongsu832 2020-07-29 09:31:30 -0400
• a611c145f4
  Add Rewrite rule to eliminate CastOp when input element type is the same as expected output element type (#237) Anh Leu 2020-07-27 11:49:14 -0500
• 32ceb6968a
  Use map when Import onnx node (#230) chentong319 2020-07-27 12:25:21 -0400
• 8af5fdeb62
  fix comment (#229) Kevin O'Brien 2020-07-24 17:03:08 -0400
• 75716baa2f
  use secret instead of token (#233) Kevin O'Brien 2020-07-24 13:31:43 -0400
• e631283c71
  ScalerOp support non-float input using CastOp (#231) Anh Leu 2020-07-24 09:57:52 -0500
• 2e8f012195
  Support krnl loop permutation (#215) Tian Jin 2020-07-24 18:19:38 +0800
• c9e3ba2d64
  Add shape inference for ScalerOp (#228) Anh Leu 2020-07-23 12:05:19 -0500
• 034f98c00c
  Shape inference for some ops in RNN models (#207) Tung D. Le 2020-07-22 23:15:56 +0900
• 1263d01968
  Add PIC property to RtMemRefUtils (#225) NathanielMcVicar 2020-07-21 23:15:25 -0700
• f43f26a79c
  modify builder (#214) chentong319 2020-07-21 22:05:18 -0400
• a58594ec81
  Revert "Emit allocs at the top of functions (#222)" (#226) Gheorghe-Teodor Bercea 2020-07-21 18:30:39 -0400
• b27e57cc4f
  Emit allocs at the top of functions (#222) Gheorghe-Teodor Bercea 2020-07-20 19:24:17 -0400
• 4b33c312d6
  Add ONNXScalerOp pattern (#220) Anh Leu 2020-07-17 10:01:30 -0500
• 13b8591af8
  add AnyMemRef (#219) chentong319 2020-07-14 11:15:06 -0400
• 0df421f7ad
  Fix importing none inputs (#217) Tung D. Le 2020-07-14 23:06:58 +0900
• d235f248e4
  Add emitjni target (#204) gongsu832 2020-07-11 01:23:13 -0400
• af75b4c75e
  Add selected MLIR options (#213) Gheorghe-Teodor Bercea 2020-07-09 18:12:57 -0400
• c557bd3444
  remove access toke, use secret instead (#212) Kevin O'Brien 2020-07-09 15:32:06 -0400
• 8848464a23
  changes for mypipeline.onnx (#202) chentong319 2020-07-08 16:52:56 -0400
• 100bfc81b4
  Bundle individual memory pools into a single memory pool (#183) Gheorghe-Teodor Bercea 2020-07-08 12:35:31 -0400
• 0a936edf79
  Remove ImportNodeReshape (#208) Tung D. Le 2020-07-08 23:27:14 +0900
• 01a4977c74
  Remove optimize_loops/return_loops op. (#200) Tian Jin 2020-07-08 12:49:15 +0800
• 07757a28ce
  Mention how to change LLVM commit level (#210) Kevin O'Brien 2020-07-07 16:43:01 -0400
• 0d375dd75e
  allow build of prereq docker image on forks (#209) Kevin O'Brien 2020-07-07 12:42:05 -0400
• 8e6642b2bc
  Update llvm version (#187) Aaron Smith 2020-07-07 06:26:00 -0700
• 8b6cf7c661
  Remove dangling constant op (#205) Tung D. Le 2020-07-07 14:47:54 +0900
• 7e05f371de
  Replace std.load/std.store by affine.load/affine.store (#180) Tung D. Le 2020-07-05 17:20:21 +0900
• 2c8f5701bd
  Lower SqueezeOp to Krnl dialect (#164) Tung D. Le 2020-07-03 17:26:41 +0900
• 4d96247327
  Set alignment attribute to AllocOp of insertAllocAndDealloc() (#195) Haruki Imai 2020-07-02 16:21:01 +0900
• 5c6d85e6f3
  README change to add info on using prebuilt Docker images (#201) Kevin O'Brien 2020-07-02 01:54:38 -0400
• cf96d635cc
  Compile cruntime with fpic (#188) gongsu832 2020-07-01 03:06:55 -0400
• e902506ee5
  Support encoding data type infomration as part of the DMR struct. (#178) Tian Jin 2020-06-30 10:58:21 +0800
• f9cb113a84
  Krnl block (#170) Tian Jin 2020-06-27 22:35:01 +0800