-
### 🐛 Describe the bug
When compiling flex_attention, any backwards operation crashes with the error `BypassFxGraphCache: Can't cache HigherOrderOperators.`.
Without compile its fine, but slow. I …
-
Failer op in [Inception_v4_vaiq_int8.default.onnx.torch.elide.mlir](https://gist.github.com/AmosLewis/c9b0933d0d1d69176d89298e7b1ff8b5#file-inception_v4_vaiq_int8-default-onnx-torch-elide-mlir)
- ht…
-
https://github.com/ROCm/AMDMIGraphX/pull/3010/files#r1648254890
With #3010 MIGraphX can fuse pointwise inputs for the Dot/conv instruction for MLIR.
It is not handling Reshapes that happen on po…
-
Hi,
I [made this fork to experiment with adding unit tests for this template and I think I was successful.](https://github.com/efferifick/mlir-standalone-template/commit/ac1eb88a9628b60498dfaac08c7…
-
python: /project/lib/Dialect/TritonGPU/IR/Dialect.cpp:52: llvm::SmallVector mlir::triton::gpu::getElemsPerThread(mlir::Attribute, llvm::ArrayRef, mlir::Type): Assertion `0 && "getElemsPerThread not im…
xueaa updated
2 months ago
-
-
After bumping llvm-project to https://github.com/llvm/llvm-project/commit/266a5a9cb9daa96c1eeaebc18e10f5a37d638734, one or more Stablehlo test(s) crash and cause the CI to timeout here:
https://githu…
-
### 1. System information
- OS Platform and Distribution: Debian GNU/Linux 10
- TensorFlow installation: pip package (Python 3.12.4)
- TensorFlow library version: 2.17.0
### 2. Code
I'm trying …
-
When you load a shared library into `mlir-cpu-runner` (in this case, the library is the HIP runtime, which is linked in by the MLIR ROCm runtime wrappers), that library's functions that have `attribut…
-
`memoryview(memref).tolist()` does not print the values as expected for some dtypes, such as `bfloat16, float8`.
Also consider move `tolist()` and `pretty_print` APIs to MLIR-TRT's python bindings.