8 Pull requests merged by 6 people

• Integrate LLVM at d78ff5f6a9c1e63a9a33ef92778edec37bf4d292

  #4180 merged May 13, 2025

• build: manually update PyTorch version

  #4179 merged May 13, 2025

• Use ceilMode from pytorch MaxPool in shlo to get correct output shape

  #4167 merged May 9, 2025

• [CI] Enable building StableHLO in CI

  #4170 merged May 8, 2025

• [ONNX] Modify QLinear* ops lowering to use common utilities

  #4178 merged May 8, 2025

• [docs] Make it more obvious that mlir_venv isn't automatically activated

  #3932 merged May 7, 2025

• Bump the github-actions group with 3 updates

  #4173 merged May 7, 2025

• [TorchToLinalg] : Lower count_nonzero.dim_IntList to Linalg-on-Tensors

  #4146 merged May 7, 2025

2 Issues opened by 2 people

• No linux prebuilt wheel for recent version of torch-mlir

  #4182 opened May 13, 2025

• Convert onnx Roialign to linalg ir error

  #4181 opened May 13, 2025

5 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• Average pooling clamped divisor should be done on all conditions where the kernel can go out of bounds

  #4144 commented on May 12, 2025 • 8 new comments

• Assertion when lowering from Torch IR for `AvgPool2d` when `kernel` is a tuple of a single int

  #3885 commented on May 8, 2025 • 0 new comments

• [ONNX] Add support for asymmetric padding for Onnx.AveragePool op

  #3923 commented on May 7, 2025 • 0 new comments

• [Torch] add aten.bilinear op decomposing

  #3931 commented on May 13, 2025 • 0 new comments

• [Torch] Add support for aten.round.decimals op

  #4166 commented on May 14, 2025 • 0 new comments