Skip to content

Train crashed #15

@whu-lyh

Description

@whu-lyh

the training log is as followings:

| Batch [2545/3811] | loss: 0.3767 |
| Batch [2546/3811] | loss: 0.3716 |
| Batch [2547/3811] | loss: 0.3690 |

the training process is stuck here and the GPU memory is enough.
Has anyone else met this before?

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