Yiakwy
Yiakwy
Hi @TamirFriedman-RecoLabs Are you working on encoder stack ? For example generate model for video, music and so on. Are you still working on this branch ? Happy to hear...
> > this doesn't work for me again, might be because I have. cc @tridao not sure how relevant this is > > The q, k, v need to be...
it may relates my local env `docker run --rm -w /app/ -v $(pwd):/app/ $image bash -c "./build-compiler.sh"` . If one want to bind `pwd` to a docker path, this may...
@zyx-billy Usually, in this workstation, we don't build anything in the physical machine, and just start a docker or using a cmake package manage system to pull dependencies in a...
@zyx-billy adding `-u $(id -u)` won't work ``` # adding -u $(id -u) won't work. One needs to register the user and annotate the user no passwd required (py39) leiw@sgjur-pod006-3:~/WorkSpace/Github/mlir-playground/cpp$...
Pytorch has decided to drop NVfuser support , see this [PR#105789](https://github.com/pytorch/pytorch/pull/105789) which later reverted by @DanilBaibak . Not sure whether they still have the plan to move forward. But it...
Hi @ppanchad-amd , I am not sure this is right place to disscuss. I have a related question, suppose I have handled a .cu file well with USE_ROCM control to...
@anilbommareddy it is generally not recommended to compile hip program with g++ (link with g++), since HIP api changes quickly. In the recent programs, if you want to compile with...
Lazy computation of partial gradients of weights with an aid of queue is really smart!. @ufotalent However, I don't believe that you need to support **sequence parallel**, a.k.a it does...
> > Hi @yxyOo: I have a few questions about total_parameters computing. Since you mentioned your experiments on llama, but I find some inconsistency: > > > > 1. llama...