How to finetune a model with some weights fronzen and some weights having different learning rate?

Closed this issue a year ago · 2 comments

Deleted user commented 2 years ago

Is there any example?

liuliu commented a year ago

See https://github.com/liuliu/swift-diffusion/blob/main/examples/lora_training/main.swift and https://github.com/liuliu/swift-diffusion/blob/main/src/LoRA.swift#L434

Deleted user commented a year ago

Thanks