Skip to content

[Feature]: Merge modeling files that remove "patches" #12713

@govind-ramnarayan

Description

@govind-ramnarayan

🚀 The feature, motivation and pitch

After merging the infrastructure changes, one additional piece of modeling infra that needs to be removed are the "patches" used to patch models beyond their HuggingFace definition to make them AutoDeploy compatible. The new sharding IR relies on custom modeling files, so we want to move entirely to these.

Alternatives

No response

Additional context

No response

Before submitting a new issue...

  • Make sure you already searched for relevant issues, and checked the documentation and examples for answers to frequently asked questions.

Metadata

Metadata

Labels

Model customization<NV>Adding support for new model architectures or variantsfeature requestNew feature or request. This includes new model, dtype, functionality support

Type

No type

Projects

Status

In review

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions