Vallabh Natu

> File "d:\data\ai\stable_diffusion\stable-diffusion\ldm\modules\encoders\modules.py", line 149, in __init__ > self.tokenizer = CLIPTokenizer.from_pretrained(version,local_files_only=True) Remove local_files_only=True so that the transformers can be downloaded from huggingface if not locally present.

Removing `--wbits 4 --groupsize 128` flags worked for me with the OPT model. I was getting same error.