Closed xerkey closed 1 month ago
@xerkey hey, i think you are passing the original model instead of the reft_model
to your trainer!
trainer = pyreft.ReftTrainerForCausalLM(
model=model, # <----- here should be reft_model
tokenizer=tokenizer, args=training_args, **data_module)
@frankaging oh my god! it's simple mistake... Thank you very much!
I have an error on trainer.trian(). Plese help me!
Error
code
Environment