Skip to content

Sudden Increase in Loss during Model Training #6

@zy-cmd

Description

@zy-cmd

Hello
I am currently working on training an RNA language model, and I have encountered an issue where the loss value suddenly spikes during the training process. It happens around 0.67 epoch. The loss suddenly increases and there is a gradient explosion. Have you ever encountered a similar problem?

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    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