-
Here's the Tom's idea from https://github.com/Mikolaj/horde-ad/issues/95#issuecomment-1503013877:
> > the transposition can't be expressed as rewriting rules, because it's stateful.
>
> It's only…
-
Nice work, this looks very promising.
I was thinking of modifying it for something I'm doing since it doesn't seem to implement return_sequences so that you can return a whole sequence of outputs? Or…
-
Hi all,
I've just found a problem through the use of a library invoking the transpose of matrix blocks. Whenever "blocks" are real, of course the "transpose" should become inoperative. However I ha…
-
BLAS installation has been a recurrent problem for users and CI:
- IRC: https://irclogs.nim-lang.org/03-03-2020.html#17:24:53
- In Nim CI due to Linux distros using either blas or cblas: https://g…
-
The following is an example of a group quantized matmul found in Vicuna (pulled from https://github.com/nod-ai/SHARK/issues/1630, closely related to the i4 IR attached in #12859).
```
#map = affine_…
-
### Feature request
MistralAI recently [released their new model](https://twitter.com/MistralAI/status/1733150512395038967), a Mixture of Experts based on [megablocks](https://github.com/stanford-fut…
-
(raised by @wacky6 Jiewei in Chromium CL review https://chromium-review.googlesource.com/c/chromium/src/+/4940628/comments/88fd209e_269f2c93)
[WebNN's matmul](https://www.w3.org/TR/webnn/#api-mlgra…
-
### What happened?
Here is the IR for a typical attention layer of bert-base-uncased HuggingFace model with 1 sentence, sequence length 128: [gist](https://gist.github.com/silvasean/304cef5cf0449f3…
-
Hello,
I'm just starting exploring the possibilities of the `xtensor-stack`, it was a bit rough at first as I had not done proper C++ in a while but I found my way around it after some time. Howeve…
-
I am working on NVGPU and NVVM dialects for Nvidia Hopper support. After having discussions at the LLVM US meeting, I want to make a list of improvements and opportunities in these dialects. I am look…
grypp updated
9 months ago