The from_pretrained method was applying LoRA twice:
1. In the constructor via lora_r parameter
2. When loading with PeftModel.from_pretrained()
Now creates model with lora_r=0 and loads LoRA weights separately.
Note: Warning about "missing adapter keys" for layers 0-11 is expected
since those layers are frozen and don't have LoRA adapters.
1bf9985def