Skip to content

KeyError when load_lora_weights in run_all.py #5

@qcy22

Description

@qcy22

I added "conv" "conv1" and "conv2" to the target_modules .Now target_modules=["to_k", "to_q", "to_v", "to_out.0", "add_k_proj", "add_v_proj","conv1","conv2","conv"] .It trained well,but when pipe.load_lora_weights in run_all.py a key error was encountered.The following picture describes the error message
640c23941449e05191e6ec15742274c
jittor=1.3.9.8
peft= 0.10.0

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