)]}'
{
  "commit": "e3f2d47cd7ae68d3c4ee68d064d1be725d5a6d55",
  "tree": "3f80db3130941381e26d3dd05c67d76f9f664a0b",
  "parents": [
    "81c8b257cbd8e5b51517e9c696210a71945f4d4b"
  ],
  "author": {
    "name": "Vivek Khandelwal",
    "email": "vivekkhandelwal1424@gmail.com",
    "time": "Wed Oct 23 18:33:01 2024 +0530"
  },
  "committer": {
    "name": "GitHub",
    "email": "noreply@github.com",
    "time": "Wed Oct 23 18:33:01 2024 +0530"
  },
  "message": "Bump torch-mlir to 140cad5 and update TorchOnnxToTorch conversion pipeline (#18867)\n\nThis commit bumps torch-mlir to\r\nhttps://github.com/llvm/torch-mlir/commit/140cad5659bb779bb1f5de1888566db5b5d21236.\r\n\r\nThis commit also replaces the `TorchOnnxToTorchPass` used to convert\r\ntorch-onnx IR to torch IR with `TorchOnnxToTorchBackendPipeline` which\r\nwas introduced here:\r\nhttps://github.com/llvm/torch-mlir/commit/fa4794dae2057876ec8ad2a6464e2668f6a2ea0c.\r\nThis pipeline consists of passes like `TorchOnnxToTorch`,\r\n`ScalarizeShapes`, `DecomposeComplexOps`, `ShapeRefinementPipeline`,\r\netc., to convert the torch-onnx IR to torch IR consistent with the other\r\nlowering paths for torch backend IR.\r\n\r\nWith the changes made in this PR, some of the tests which were failing\r\nearlier during compilation now passes compilation, some of them even\r\npasses inference, while there are some tests which were passing earlier\r\nare now failing. The list is as follows:\r\n- For\r\ntests/external/iree-test-suites/onnx_ops/onnx_ops_cpu_llvm_sync.json:\r\n  a.) Earlier failing compilation, now passing inference:\r\n    - onnx/node/generated/test_convtranspose_kernel_shape\r\n    - onnx/node/generated/test_einsum_sum\r\n\r\nb.) Earlier failing compilation, now passing compilation but failing\r\ninference\r\n    - onnx/node/generated/test_convtranspose_output_shape\r\n   \r\n  c.) Earlier passing but now failing inference\r\n    - onnx/node/generated/test_scan_sum\r\n    - onnx/node/generated/test_scan9_sum\r\n\r\n- For\r\ntests/external/iree-test-suites/onnx_ops/onnx_ops_gpu_rocm_rdna3.json:\r\n  a.) Earlier failing compilation, now passing inference:\r\n    - onnx/node/generated/test_convtranspose_kernel_shape\r\n    - onnx/node/generated/test_einsum_sum\r\n\r\nb.) Earlier failing compilation, now passing compilation but failing\r\ninference\r\n    - onnx/node/generated/test_convtranspose_output_shape\r\n   \r\n  c.) Earlier passing but now failing inference\r\n    - onnx/node/generated/test_scan_sum\r\n    - onnx/node/generated/test_scan9_sum\r\n    - onnx/node/generated/test_slice_default_axes\r\n \r\n- For tests/external/iree-test-suites/onnx_ops/onnx_ops_gpu_vulkan.json:\r\n  a.) Earlier failing compilation, now passing inference:\r\n    - onnx/node/generated/test_convtranspose_kernel_shape\r\n\r\nb.) Earlier failing compilation, now passing compilation but failing\r\ninference\r\n    - onnx/node/generated/test_convtranspose_output_shape\r\n    - onnx/node/generated/test_einsum_sum\r\n\r\n  c.) Earlier passing but now failing inference\r\n    - onnx/node/generated/test_scan_sum\r\n    - onnx/node/generated/test_scan9_sum\r\n \r\nThis commit\r\nhttps://github.com/llvm/torch-mlir/commit/55ff110dc29cab7e2495ccdbec9a60512c29c665\r\nis expected to fix the newly introduced failures which will be included\r\nin IREE in the next Torch-MLIR bump.\r\n\r\n---------\r\n\r\nSigned-off-by: Vivek Khandelwal \u003cvivekkhandelwal1424@gmail.com\u003e",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "b9a497f12550a5fd2833efa2ffdbf0b1fb7b3825",
      "old_mode": 33188,
      "old_path": "compiler/plugins/input/Torch/PluginRegistration.cpp",
      "new_id": "6f79686ad267ec8c4918248817e6840064bb08fc",
      "new_mode": 33188,
      "new_path": "compiler/plugins/input/Torch/PluginRegistration.cpp"
    },
    {
      "type": "modify",
      "old_id": "0bbd604384f8eda3cec35daf3f4492ec5b6f2954",
      "old_mode": 33188,
      "old_path": "tests/external/iree-test-suites/onnx_ops/onnx_ops_cpu_llvm_sync.json",
      "new_id": "a025431d7af4d0290daa79fe3f48fea97b64fe1f",
      "new_mode": 33188,
      "new_path": "tests/external/iree-test-suites/onnx_ops/onnx_ops_cpu_llvm_sync.json"
    },
    {
      "type": "modify",
      "old_id": "9901e017948cacca645f5c9740b475645f50eedc",
      "old_mode": 33188,
      "old_path": "tests/external/iree-test-suites/onnx_ops/onnx_ops_gpu_rocm_rdna3.json",
      "new_id": "79cc5a9a4addcbae72e9be1c04584352f67fccf3",
      "new_mode": 33188,
      "new_path": "tests/external/iree-test-suites/onnx_ops/onnx_ops_gpu_rocm_rdna3.json"
    },
    {
      "type": "modify",
      "old_id": "e6d9a4a4e2014ed201d5bfd8bf2f911a1ff7b508",
      "old_mode": 33188,
      "old_path": "tests/external/iree-test-suites/onnx_ops/onnx_ops_gpu_vulkan.json",
      "new_id": "8c31c26a421a159e0ba908363f93bdf198ee8e0a",
      "new_mode": 33188,
      "new_path": "tests/external/iree-test-suites/onnx_ops/onnx_ops_gpu_vulkan.json"
    },
    {
      "type": "modify",
      "old_id": "45bb17ebfe5e9cdcfd2cfabf850d9dec7127c5ab",
      "old_mode": 57344,
      "old_path": "third_party/torch-mlir",
      "new_id": "140cad5659bb779bb1f5de1888566db5b5d21236",
      "new_mode": 57344,
      "new_path": "third_party/torch-mlir"
    }
  ]
}