Skip to content

decouple IDX format from raw tensor #42

Description

@lgarithm

The IDX format only defined encodings for: u8, i8, i16, i32, f32, f64.
The new encoding format should support more scalar data types, and should be extendable.

Activity

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

Metadata

Metadata

Assignees

Labels

No labels
No labels

Type

No type

Projects

No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions