BLIP captioning broken

[nathankrueger]

Using Python 3.10.9 on Windows 11

Current Working Directory is: C:\Users\natek\Apps\Khoya\kohya_ss load images from C:\Users\natek\Downloads\test_blip found 3 images. loading BLIP caption: https://storage.googleapis.com/sfr-vision-language-research/BLIP/models/model_large_caption.pth Traceback (most recent call last): File "C:\Users\natek\Apps\Khoya\kohya_ss\finetune\make_captions.py", line 202, in main(args) File "C:\Users\natek\Apps\Khoya\kohya_ss\finetune\make_captions.py", line 88, in main model = blip_decoder(pretrained=args.caption_weights, image_size=IMAGE_SIZE, vit="large", med_config="./blip/med_config.json") File "C:\Users\natek\Apps\Khoya\kohya_ss\finetune\blip\blip.py", line 175, in blip_decoder model = BLIP_Decoder(*kwargs) File "C:\Users\natek\Apps\Khoya\kohya_ss\finetune\blip\blip.py", line 98, in init self.tokenizer = init_tokenizer() File "C:\Users\natek\Apps\Khoya\kohya_ss\finetune\blip\blip.py", line 189, in init_tokenizer tokenizer = BertTokenizer.from_pretrained('bert-base-uncased') File "C:\Users\natek\Apps\Khoya\kohya_ss\venv\lib\site-packages\transformers\tokenization_utils_base.py", line 2028, in from_pretrained return cls._from_pretrained( File "C:\Users\natek\Apps\Khoya\kohya_ss\venv\lib\site-packages\transformers\tokenization_utils_base.py", line 2260, in _from_pretrained tokenizer = cls(init_inputs, **init_kwargs) File "C:\Users\natek\Apps\Khoya\kohya_ss\venv\lib\site-packages\transformers\models\bert\tokenization_bert.py", line 199, in init if not os.path.isfile(vocab_file): File "C:\Users\natek.pyenv\pyenv-win\versions\3.10.9\lib\genericpath.py", line 30, in isfile st = os.stat(path) TypeError: stat: path should be string, bytes, os.PathLike or integer, not NoneType 18:58:26-939255 INFO ...captioning done

[bmaltais]

This is an issue that should be directly to kohya's sd-script repo. I only copy his script files and this is not one I maintain.