No merges.txt file with text generation webui
#7
by
ranguna
- opened
When loading this model through web text generation webui, I'm getting the following error:
2023-07-09 00:52:17 ERROR:Failed to load the model.
Traceback (most recent call last):
File "$HOME/projects/ai/text-generation-webui/server.py", line 68, in load_model_wrapper
shared.model, shared.tokenizer = load_model(shared.model_name, loader)
File "$HOME/projects/ai/text-generation-webui/modules/models.py", line 87, in load_model
tokenizer = load_tokenizer(model_name, model)
File "$HOME/projects/ai/text-generation-webui/modules/models.py", line 104, in load_tokenizer
tokenizer = AutoTokenizer.from_pretrained(
File "$HOME/.miniconda3/envs/textgen/lib/python3.10/site-packages/transformers/models/auto/tokenization_auto.py", line 691, in from_pretrained
return tokenizer_class.from_pretrained(pretrained_model_name_or_path, *inputs, **kwargs)
File "$HOME/.miniconda3/envs/textgen/lib/python3.10/site-packages/transformers/tokenization_utils_base.py", line 1825, in from_pretrained
return cls._from_pretrained(
File "$HOME/.miniconda3/envs/textgen/lib/python3.10/site-packages/transformers/tokenization_utils_base.py", line 1989, in _from_pretrained
tokenizer = cls(*init_inputs, **init_kwargs)
File "$HOME/.miniconda3/envs/textgen/lib/python3.10/site-packages/transformers/models/gpt2/tokenization_gpt2.py", line 195, in __init__
with open(merges_file, encoding="utf-8") as merges_handle:
TypeError: expected str, bytes or os.PathLike object, not NoneType
Logging call arguments passed to tokenization_utils_base.py
, we can see that the merges file is indeed missing:
{'add_prefix_space': False, 'additional_special_tokens': ['<|endoftext|>', '<fim_prefix>', '<fim_middle>', '<fim_suffix>', '<fim_pad>', '<filename>', '<gh_stars>', '<issue_start>', '<issue_comment>', '<issue_closed>', '<jupyter_start>', '<jupyter_text>', '<jupyter_code>', '<jupyter_output>', '<empty_output>', '<commit_before>', '<commit_msg>', '<commit_after>', '<reponame>'], 'bos_token': '<|endoftext|>', 'clean_up_tokenization_spaces': True, 'eos_token': '<|endoftext|>', 'model_max_length': 1000000000000000019884624838656, 'unk_token': '<|endoftext|>', 'vocab_size': 49152, 'vocab_file': 'models/TheBloke_starchat-beta-GPTQ/vocab.json', 'merges_file': None, 'special_tokens_map_file': 'models/TheBloke_starchat-beta-GPTQ/special_tokens_map.json', 'name_or_path': 'models/TheBloke_starchat-beta-GPTQ'}
Is this file missing from this model or is this something with my local setup ?
Sorry that was my fault. merges.txt is uploaded now. Trigger another download of the repo and it will download the missing file and then it should work
No worries
It's working perfectly now, thanks for the awesome work!
Btw, it seems the merges.txt file is also missing in this model: https://huggingface.co./TheBloke/starcoderplus-GPTQ
Would you like me to open a discussion there ?
OK thanks, that's fixed now too.
Amazing. Thanks again!
ranguna
changed discussion status to
closed