Skip to content

Neuralprophet 0.9 error #1689

@Aaron1993

Description

@Aaron1993

When I use neuralprophet=0.9 and python=3.10 to predict, the following code raises an error CUDA error: out of memory.
m = neuralprophet()
m.fit(df, freq='D')

However, when I change the code as
m = neuralprophet(train_config={'accelerator0: 'cpu''})
m.fit(df, freq='D')
There is raised another error Failed to fit model, Weights_only load failed. unsupported global: GLOBAL neuralprophet.configure.ConfigSeasonality was not an allowed global by default

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions