Skip to content

Change default Pytorch model for STX/KRK#331

Open
danielholanda wants to merge 4 commits into
mainfrom
dholanda/pytorch_core_model
Open

Change default Pytorch model for STX/KRK#331
danielholanda wants to merge 4 commits into
mainfrom
dholanda/pytorch_core_model

Conversation

@danielholanda
Copy link
Copy Markdown
Collaborator

Description

Change default Pytorch model for STX/KRK.

@sreeram-11
Copy link
Copy Markdown
Collaborator

Awesome! Looks great

@danielholanda danielholanda marked this pull request as ready for review June 2, 2026 22:54
@danielholanda
Copy link
Copy Markdown
Collaborator Author

@adamlam2-amd
Can you please also review?
This PR enables STX/KRK with a smaller model and I just want to make sure we are not introducing any bugs to the halo path that should*** be unchanged.

### Installing Additional Dependencies

<!-- @device:halo,halo_box -->
<!-- @var:id=hf_model value="openai/gpt-oss-20b" -->
Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

does this install for halo_box? the model is already pre-installed for halo_box, fyi.

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

if it just sets the var then nevermind haha

@adamlam2-amd
Copy link
Copy Markdown
Collaborator

as well, this may be something to think about for the future, but the RAM that comes with the processors is not fixed. For example, a strix halo can come with only 32GB RAM, which would not work with the larger models.

@adamlam2-amd
Copy link
Copy Markdown
Collaborator

i didn't test it myself; please make sure you do before merging. thanks

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants