Windows AMD DirectML support

[PixifyAI]

Could we get an implementation (install_win_AMD.cmd) of DirectML for Windows as in auto1111 https://github.com/lshqqytiger/stable-diffusion-webui-directml

[Woolverine94]

Hello @PixifyAI ,

First of all, thanks for your contribution.

From what I understand of DirectML and what I know of biniou, it doesn't seems to be an acceleration that can be integrated at this time (see here : pytorch 2.1.0 is the only pytorch version compatible with all pipelines used in biniou). More of this, I have no way to test it as I don't have access to AMD GPU + Windows environment.

However, this is a very good suggestion, and I will try to keep an eye on this.

[PixifyAI]

I have 6950 xt and Ryzen 9 5950X with 32 gig ram an 19T of HD space for all the models you could want to use. If you want to test anything im down to help out, just let me know.

…

On Tue, Mar 5, 2024 at 3:56 PM Woolverine94 ***@***.***> wrote: Hello PixifiAI, First of all, thanks for your contribution. From what I understand of DirectML and what I know of biniou, it doesn't seems to be an acceleration that can be integrated at this time (see here <microsoft/DirectML#415 (comment)> : pytorch 2.1.0 is the only pytorch version compatible with all pipelines used in biniou). More of this, I have no way to test it as I don't have access to AMD GPU + Windows environment. However, this is a very good suggestion, and I will try to keep an eye on this. — Reply to this email directly, view it on GitHub <#15 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BGWIGWGWX4R7OKFK4IZPS7LYWY5PVAVCNFSM6AAAAABEHWRHCGVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4DMOBVHE4DG> . You are receiving this because you authored the thread.Message ID: ***@***.***>