Skip to content

persimmon crashes with CUDA: assertion failure ggml_is_contiguous(src0) #5823

Open
@cebtenzzre

Description

@cebtenzzre

Attempting to run a persimmon model with the CUDA backend fails an assertion in ggml_cuda_rope: ggml_is_contiguous(src0)

ref #5668 (comment)

Metadata

Metadata

Assignees

No one assigned

    Labels

    bugSomething isn't workingmodelModel specific

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions