llvm/torch-mlir
The Torch-MLIR project aims to provide first class support from the PyTorch ecosystem to the MLIR ecosystem.
C++NOASSERTION
Issues
- 5
[ONNX] Add support for onnx.Resize with nearest_mode "round_prefer_floor" and "round_prefer_ceil"
#3420 opened by flemairen6 - 1
Failed to lower "torch.aten.convolution" to linalg
#3429 opened by josel-amd - 0
Error lowering `index_put_`: 'tm_tensor.scatter' op mismatch in shape of indices and update value at dim#0
#3433 opened by ScottTodd - 1
- 0
- 0
- 0
Failed to lower "torch.aten.sum.dim_IntList" to linalg
#3423 opened by josel-amd - 0
- 0
Failed to lower torch.aten.slice.Tensor to linalg
#3417 opened by josel-amd - 11
[ONNX] Systematically expand ONNX functions before conversion to Torch to avoid needing bespoke conversions
#3384 opened by andfau-amd - 0
Crash with TorchBackendToLinalgOnTensorsBackendPipeline
#3352 opened by zjgarvey - 2
torchscript_stablehlo_backend_tinybert.py failed to lower Torch Backend IR -> StableHLO
#3411 opened by FruitClover - 2
- 0
onnx.MatMul version 1 not lowered (but it's essentially the same as version 13)
#3324 opened by mgehre-amd - 1
migrate to from member `x.cast<T>()` to `mlir::cast<T>(x)`
#3299 opened by bjacob - 7
- 11
Some issues with migrating to one-shot-bufferize
#3399 opened by zjgarvey - 4
- 0
unsupported aten.meshgrid operator
#3366 opened by zhangdong-user - 0
link out of date
#3356 opened by ai-mannamalai - 0
onnx.Transpose version 1 not lowered (but it's essentially the same as version 13)
#3290 opened by mgehre-amd - 3
Missing dtype support for OnnxToTorch
#3255 opened by jinchen62 - 4
Update version pin to onnx 1.16
#3310 opened by stellaraccident - 1
[ONNX] Unsqueeze -> Torch -> Linalg issue
#3101 opened by zjgarvey - 0
QuantizedMLP currently failing in CI
#3298 opened by zjgarvey - 0
- 0
`NotSupportedError` error when executing Readme example "TorchScript ResNet18"
#3284 opened by trirpi - 5
[ONNX] onnx.Loop
#3136 opened by renxida - 0
iree-compiler failed to compile gpt2
#3267 opened by Peter9606 - 4
Strange issue related to `pyproject.toml` when building release wheels
#3258 opened by sjain-stanford - 1
Compilation of Quantized Model Failed
#3220 opened by alexsifivetw - 1
- 2
`make_simple_dynamo_backend` does not work for single argument models.
#3204 opened by ttrouwen-dmatrix - 2
[ONNX] MatMulInteger Conversion Fails For Arg Rank != 2
#3137 opened by zjgarvey - 0
Build failed on x86 MacOS
#3191 opened by lifengxiang1025 - 0
[e2e testing] make e2e_testing/main.py print ir / more informative error messages
#3172 opened by renxida - 1
[TORCH] `AtenFloorDivideOp` truncates instead of flooring which doesn't match `torch.floor_divide()` .
#3142 opened by IanWood1 - 0
[TORCH] add support for torch.aten.floor_divide.Scalar
#3143 opened by IanWood1 - 0
Slice Folder Fails Result Type Match Assert
#3153 opened by zjgarvey - 0
- 1
- 0
- 5
[RFC] Mixed Quantized Types in Matmul Not Supported
#3090 opened by zjgarvey - 0
[MLIR][TORCH] Onnx.MaxPool fails on dynamic sizes
#3135 opened by IanWood1 - 5
- 1
LTC build error after PT bump: multiple definition of `torch::lazy::compute_shape_arange`
#3120 opened by sjain-stanford - 4
FX Importer error: `Could not deduce type from value info: tensor_meta=None, val=s0, sparsity=None`
#3118 opened by sjain-stanford - 9
Regression in onnx.PRelu
#3092 opened by stellaraccident - 0
torch_mlir.compile failing on RetinaNET
#3096 opened by afanasyev-ilya - 1
AtenSelectInt results in `failed to legalize operation 'arith.extui' that was explicitly marked illegal`
#3077 opened by renxida