Skip to content

Support for machine learning dtypes #372

@justinchuby

Description

@justinchuby

I am building a reference runtime for ONNX using this library, which needs to support data types commonly used in machine learning, e.g. bfloat16, float8*, float4, and lower bit ints like (u)int4 and (u)int2. These types exist in frameworks like jax, pytorch and numpy (via the ml_dtypes package). Is there a plan to support these datatypes?

Metadata

Metadata

Assignees

No one assigned

    Labels

    questionFurther information is requested

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions