AssertionError: duplicated lora name when loading Lora via create_network_from_weights with diffusers 0.15

Hello, Official release for diffusers 0.15 just came out and loading Lora from create_network_from_weights in lora.py now results in the following error:

```
assert lora.lora_name not in names, f"duplicated lora name: {lora.lora_name}"
AssertionError: duplicated lora name: lora_unet_down_blocks_0_attentions_0_transformer_blocks_0_attn1_to_q
```

Here is the code that worked on diffusers 0.14:

```
def apply_lora(pipe, lora_path, weight:float = 1.0):
from safetensors.torch import load_file
from utils.lora import create_network_from_weights
import torch

vae = pipe.vae
text_encoder = pipe.text_encoder
unet = pipe.unet

lora_path = "./lora/" + lora_path
sd = load_file(lora_path)
lora_network, sd = create_network_from_weights(weight, None, vae, text_encoder, unet, sd)
lora_network.apply_to(text_encoder, unet)
lora_network.load_state_dict(sd)
lora_network.to("cuda", dtype=torch.float16)
```

Please advice on how to load safetensor Lora Models on Diffusers 0.15 pipelines.

Thanks!

Start in lora.py at create_network_from_weights and reproduce the duplicated-name assertion with the provided apply_lora example and a Diffusers 0.15 pipeline. Compare the relevant loading behavior against Diffusers 0.14. Done means the reported LoRA model loads without the duplicate-name assertion on the newer pipeline.

Written by the indexing model from the issue text.