Code

Verbose#3937

Code

Draft

rahuls-cerebras wants to merge 47 commits intomainllvm/torch-mlir:mainfrom rahuls/verbose_modellvm/torch-mlir:rahuls/verbose_modeCopy head branch name to clipboard

+1,597-6,785

Commits

Commits on Jan 3, 2025

• Revert "Fix output size computation for MaxPool2D for ceil_model = true. (#3890)"

  Show description for 9c5055c
  rahuls-cerebras

  committed
  9c5055c

• Revert "build: manually update PyTorch version (#3919)"

  Show description for 23453db
  rahuls-cerebras

  committed
  23453db

• Revert "[torch-mlir][doc] remove MPACT as example (#3930)"

  Show description for f70994d
  rahuls-cerebras

  committed
  f70994d

• Revert "[TOSA] Add legalization for torch.aten.unfold (#3922)"

  Show description for 9e6493e
  rahuls-cerebras

  committed
  9e6493e

• Revert "[torch-mlir] add MPACT as an example torch-mlir based compiler (#3928)"

  Show description for afae33e
  rahuls-cerebras

  committed
  afae33e

• Revert "[Torch] Add decomposition for 1d torch.nonzero (#3876)"

  Show description for 31b46a3
  rahuls-cerebras

  committed
  31b46a3

• Revert "[torch] Update torch.bmm to use accumulator type (#3924)"

  Show description for 327e6f2
  rahuls-cerebras

  committed
  327e6f2

• Revert "Add attributes support for onnx.nms (#3920)"

  Show description for 6acf852
  rahuls-cerebras

  committed
  6acf852

• Revert "[torch-mlir][sparse] register sparse tensor dialect for all rewriting (#3918)"

  Show description for 9c4dd02
  rahuls-cerebras

  committed
  9c4dd02

• Revert "[MLIR][TORCH] Add support for 1-d group convolution (#3904)"

  Show description for f118b62
  rahuls-cerebras

  committed
  f118b62

• Revert "[ONNX] Fix nonzero output type difference between onnx and torch (#3916)"

  Show description for c6c429f
  rahuls-cerebras

  committed
  c6c429f

• Revert "[TorchToTosa] Refactoring to separate construction of legal/illegal ops and conversion patterns. (#3759)"

  Show description for 51ec34c
  rahuls-cerebras

  committed
  51ec34c

• Revert "[MLIR][TORCH] Add aten.special.expm1 op lowering (#3878)"

  Show description for 5de1e25
  rahuls-cerebras

  committed
  5de1e25

• Revert "Replace unmaintained create-release action (#3905)"

  Show description for ab0819f
  rahuls-cerebras

  committed
  ab0819f

• Revert "Replace ubuntu-latest with specific version (#3906)"

  Show description for 811e887
  rahuls-cerebras

  committed
  811e887

• Revert "Pin and update actions (#3907)"

  Show description for edb27bd
  rahuls-cerebras

  committed
  edb27bd

• Revert "build: manually update PyTorch version (#3896)"

  Show description for 15e2d4a
  rahuls-cerebras

  committed
  15e2d4a

• Revert "[TOSA] Add some more mixed dtype handling (#3909)"

  Show description for 671656a
  rahuls-cerebras

  committed
  671656a

• Revert "[onnx][importer] Merge character sanitization into regex (#3914)"

  Show description for 906a0c3
  rahuls-cerebras

  committed
  906a0c3

• Revert "[onnx][importer] Sanitize '-' characters in TensorProto names (#3901)"

  Show description for 8dedb73
  rahuls-cerebras

  committed
  8dedb73

• Revert "[CI] Disable LTC build from CI (#3911)"

  Show description for c2cd79f
  rahuls-cerebras

  committed
  c2cd79f

• Revert "Add support for the padding variations of conv op (#3883)"

  Show description for f889645
  rahuls-cerebras

  committed
  f889645

• Revert "[TOSA] Add logit, log1p, log10 and add promote type to unary fponly ops (#3900)"

  Show description for 20102b7
  rahuls-cerebras

  committed
  20102b7

• Revert "[TOSA] Add upsample_nearest2d, split_dim, outer, GELU tanh mode and misc (#3886)"

  Show description for 1ae33cb
  rahuls-cerebras

  committed
  1ae33cb

• Revert "Support bf16 on aten.uniform lowering (#3895)"

  Show description for 60710ca
  rahuls-cerebras

  committed
  60710ca

• Revert "aten.pow / onnx.Pow: Fix (float,int) / (int, float) accuracy (#3894)"

  Show description for 2a6b9f6
  rahuls-cerebras

  committed
  2a6b9f6

• Revert "[TorchToLinalg] Add aten.fft_rfft and lowering (#3857)"

  Show description for 6fa8e82
  rahuls-cerebras

  committed
  6fa8e82

• Revert "[Torch Dialect] Emit torch.aten.mul.float_int, add folder and conversion to Arith. (#3750)"

  Show description for 20c65b6
  rahuls-cerebras

  committed
  20c65b6

• Revert "Support NMS op lowering (#3871)"

  Show description for 7fe2fd6
  rahuls-cerebras

  committed
  7fe2fd6

• Revert "Support stash_type attribute for onnx.LayerNormalization (#3888)"

  Show description for 4ec60e9
  rahuls-cerebras

  committed
  4ec60e9

• Revert "Add TorchToArith rewrite pattern for AtenGtFloatOp (#3892)"

  Show description for 8ad6d26
  rahuls-cerebras

  committed
  8ad6d26

• Revert "[Torch] Address unnecessary dynamic shapes in argmax decomposition (#3889)"

  Show description for 1a43203
  rahuls-cerebras

  committed
  1a43203

• Revert "convert to double before float materialization in scalarize shapes (#3887)"

  Show description for 9f0a407
  rahuls-cerebras

  committed
  9f0a407

• Revert "[Torch Dialect] Add torch.aten.mul.int_float (required to simplify shape calculation of upsample_nearest2d) (#3764)"

  Show description for ec38cc9
  rahuls-cerebras

  committed
  ec38cc9

• Revert "[TorchToStablehlo] support l1_loss, deg2rad, logit (#3865)"

  Show description for 1b6cd0f
  rahuls-cerebras

  committed
  1b6cd0f

• Revert "[Torch] support aten.column_stack (#3867)"

  Show description for e8c94b2
  rahuls-cerebras

  committed
  e8c94b2

• Revert "[TOSA] Add reflection and replication pad lowering (#3874)"

  Show description for 37feff7
  rahuls-cerebras

  committed
  37feff7

• Revert "[TorchToLinalg] Use linalg.transpose instead of generic in permuteTensor (#3872)"

  Show description for f627b06
  rahuls-cerebras

  committed
  f627b06

• Revert "Fix a bug for large models in onnx importer. (#3875)"

  Show description for 1b3bfd3
  rahuls-cerebras

  committed
  1b3bfd3

• Revert "Bump LLVM to 813f7c3 (#3873)"

  Show description for 96ffb3b
  rahuls-cerebras

  committed
  96ffb3b

• Revert "[ONNX] Remove kernel shape and weight shape equivalence check from Onnx.Conv lowering (#3869)"

  Show description for 469a7b1
  rahuls-cerebras

  committed
  469a7b1

• Revert "[Torch Dialect] Allow simplification of shape calculations of aten.tile, col2im, aten.stft (#3785)"

  Show description for 1d3c7c6
  rahuls-cerebras

  committed
  1d3c7c6

• Revert "[ONNX] rework some reduction op lowerings (#3870)"

  Show description for ca77a5a
  rahuls-cerebras

  committed
  ca77a5a

• Revert "Support default padding case for tosa::AvgPool in the presence of count_include_pad (#3868)"

  Show description for c1c0524
  rahuls-cerebras

  committed
  c1c0524

• Revert "Add Scalarization Patterns for AtenToDtypeOp, AtenNegOp, AtenRemainderTensorOp (#3861)"

  Show description for 612ccc3
  rahuls-cerebras

  committed
  612ccc3

• Revert "OnnxToTorch bicubic interpolation (#3802)"

  Show description for e7389de
  rahuls-cerebras

  committed
  e7389de

• Revert old pytorch

  Show description for 7a5f8ad
  rahuls-cerebras

  committed
  7a5f8ad