Skip to content

Tracking issue for SPHINX quantization & other memory issues #114

@linziyi96

Description

@linziyi96

We have currently received several requests (#112, #110, #97) to run the SPHINX inference on GPUs with smaller memory. We also believe that fitting it into the 24GB memory bar benefits a broad range of users who would like to run the model locally on commodity GPUs like 3090 or 4090.

With the latest update #113, we should see NF4 quantization running fine on SPHINX without errors (i.e., resolving #97). The memory usage is a bit less than 23GB, and it should now fit into a single 24GB GPU (3090, 4090 or A5000) even with ECC turned on

image

We are still doing a complete benchmark of this quantized model and will update the latest information under this issue. Meanwhile, any question is also welcomed :)

Metadata

Metadata

Assignees

No one assigned

    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