the E is for experimental. lora doesnt work well without using certain prompts. not sure if its over trained or not trained enough. Don't use the base model to make images. has some errors here and there.