-
### What happened?
- Architecture: AMD EPYC 7H12 64-Core Processor | x86_64
I am currently trying to compile the TF model to mlir and then use the IREE compiler to generate the executa…
-
Likely culprit: https://github.com/iree-org/iree/pull/11285
I'm noticing that when I make changes to compiler code and rebuild `iree-test-deps`, the `.vmfb` test files are not regenerated. I think …
-
I found this while compiling the IREE _compiler_ through Emscripten to WebAssembly (see [notes](https://gist.github.com/ScottTodd/7898981998cefb60902d4fbc8a471ccf)).
When cross-compiling the IREE _…
-
In runtime/bindings/python/vm.cc:
```
VmContext VmContext::Create(VmInstance* instance,
std::optional& modules) {
IREE_TRACE_SCOPE_NAMED("VmContext::Create");
ir…
-
### What happened?
For [UNET model](https://storage.googleapis.com/shark_tank/SDXL/mlir/unet_1_77_1024_1024_fp16_stable-diffusion-xl-base-1.mlir), the execution is extremely slow (~500/200+ sec in va…
-
# Single-node ML Runtime Foundation | Lei.Chat()
Previous blog posts overviewed the MLIR dialect hierarchy for kernel code
generation (CodeGen) and zoomed in on the
Linalg and Vector dialects among t…
-
### What happened?
I do not see GPU profile option with `./iree-tracy-profiler 1.tracy`.
 -> tensor…
-
Some of them are done on CUDA path, and the rest need to be revisited.
https://github.com/iree-org/iree/blob/1fcb89d36e1bc0d8c7cc2818b1025f182bfc4a75/tests/e2e/tensor_ops/BUILD.bazel#L76-L102
We…
-
# Leaving Google | Lei.Chat()
Time flies—almost 9 years have passed since I joined Google.
Now the time has come for me to leave and move on.
While here, I’m super lucky to mostly work on open source…