-
## Summary
Script:
----------
from concrete import fhe
def max_two_numbers(a, b):
return (a + b + abs(a - b)) // 2
def max_in_list(numbers):
max_value = numbers[0]
for num in n…
-
https://github.com/nod-ai/e2eshark-reports/blob/main/2024-06-24/onnx_reports/statusreport.md
input ir: [mit-b0.default.pytorch.linalg.elide.mlir](https://gist.github.com/AmosLewis/33187568cf09d47b3d9…
-
After #1545 designs with loops that have less iterations than the unroll factor fail silently.
It was caught by one of @eddierichter-amd designs.
The example [here](https://github.com/Xilinx/ml…
-
I believe the command to compile to native program is outdated in the `##From MLIR to native binary` section in `README.md`.
```sh
# compile natively
"$MLIR_SYS_180_PREFIX"/bin/clang program.ll -…
-
#939 removed the lines of the README regarding being able to build TableGen-derived documentation. The issue is that the `mlir-doc` target is broken at the moment.
We should investigate the issue (…
-
Such linalg.generic can not be vetorized:
```
// mlir-opt test.mlir -transform-interpreter
#map = affine_map (d0, d1, d2, d3)>
#map1 = affine_map (d0, d1, d2, 0)>
module {
func.func @generic_1…
-
The following tests does not test any stablehlo ops other than `stablehlo.constant`. From the name It seems they are intended to test some other ops but skipped somehow. Let explore fixing (delete or …
-
-
This issue has two parts.
The first part is to fuse reductions(including split reductions) with MLIR, including any pointwise.
The second part is to use multiple outputs when fusing, to fuse eve…
-
```
def export_mlir(model, dummy_input):
from torch_mlir import torchscript
module = torchscript.compile(model, dummy_input, output_type="tosa")
print("TOSA OutputType\n")
with …