jamiemoller

ps. love the work @mudler

it should be noted 1 - the documentation for rocm for some reason indicates `make BUILD_TYPE=hipblas GPU_TARGETS=gfx1030` ... there is no `build` arg 2 - stablediffusion is the hardest thing...

it should be noted that while i do see models load onto the card whenever there is an api call and there are computations being performed pushing the card to...

i have a feeling that it will be better to start from [here](https://hub.docker.com/layers/rocm/pytorch/rocm6.0_ubuntu20.04_py3.9_pytorch_2.1.1/images/sha256-118f9a292165491464f839ba144a712b2a60a15d666f67f25b10d3627da27d8b?context=explore) (or something) for amd builds now that 2.8 is on the ubu22.04

> > Im trying to work on the hipblas version but I am confused on where the Dockerfiles are located that are used to generate the latest images such as...

> Im trying to work on the hipblas version but I am confused on where the Dockerfiles are located that are used to generate the latest images such as "quay.io/go-skynet/local-ai:master-hipblas"...

> > newer does not equal better, this said, x.x.Y versions of Y variation are usually hotfixes and usually only apply to some very specific edge cases, can you clarify...

@bunder2015 when you say 'it' do you mean the container or the 'stablediffusion' backend. also would you mind listing any of the aio defined models and if they offload to...

This is the same sort of issue that I have had for a while, this issue is adjacent to https://github.com/mudler/LocalAI/issues/1592 - Limited success building on metal - it will build...

I'm gonna have a stab at this soon if i get time