Skip to content

how to finetune? #2

@haooxia

Description

@haooxia

I want to know if I need to rebuild the data set when finetune?
I use 20 fonts in pretraining, and use 1 font in finetuning, but it report an error:

InvalidArgumentError (see above for traceback): Restoring from checkpoint failed. This is most likely due to a mismatch between the current graph and the graph from the checkpoint. Please ensure that you have not altered the graph expected based on the checkpoint. Original error:

Assign requires shapes of both tensors to match. lhs shape= [900,1] rhs shape= [900,20]
         [[node save/Assign_130 (defined at main.py:197) ]]

Do you know how to fix it?

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