ValueError: Tokenizer class ChatGLMTokenizer does not exist or is not currently imported.

[phalexo]

(FinGPT) developer@ai:~/PROJECTS/FinGPT/fingpt/FinGPT_sentiment/instruct-FinGPT$ python ./inference/batchbot_torch.py --path /opt/data/data/THUDM/chatglm2-6b --max_new_tokens 16 Traceback (most recent call last): File "/home/developer/PROJECTS/FinGPT/fingpt/FinGPT_sentiment/instruct-FinGPT/./inference/batchbot_torch.py", line 147, in main(args) File "/home/developer/PROJECTS/FinGPT/fingpt/FinGPT_sentiment/instruct-FinGPT/./inference/batchbot_torch.py", line 93, in main generator = get_generator(args.path) File "/home/developer/PROJECTS/FinGPT/fingpt/FinGPT_sentiment/instruct-FinGPT/./inference/batchbot_torch.py", line 53, in get_generator tokenizer = AutoTokenizer.from_pretrained(path, fast_tokenizer=True) File "/home/developer/mambaforge/envs/FinGPT/lib/python3.10/site-packages/transformers/models/auto/tokenization_auto.py", line 748, in from_pretrained raise ValueError( ValueError: Tokenizer class ChatGLMTokenizer does not exist or is not currently imported. (FinGPT) developer@ai:~/PROJECTS/FinGPT/fingpt/FinGPT_sentiment/instruct-FinGPT$

[phalexo]

After lots of fiddling, now it does this:

File "/home/developer/PROJECTS/FinGPT/fingpt/FinGPT-v3/benchmark/alexotest.py", line 9, in tokenizer = AutoTokenizer.from_pretrained(base_model, trust_remote_code=True) File "/home/developer/mambaforge/envs/CodeLlama/lib/python3.10/site-packages/transformers/models/auto/tokenization_auto.py", line 738, in from_pretrained return tokenizer_class.from_pretrained(pretrained_model_name_or_path, *inputs, **kwargs) File "/home/developer/mambaforge/envs/CodeLlama/lib/python3.10/site-packages/transformers/tokenization_utils_base.py", line 2042, in from_pretrained return cls._from_pretrained( File "/home/developer/mambaforge/envs/CodeLlama/lib/python3.10/site-packages/transformers/tokenization_utils_base.py", line 2253, in _from_pretrained tokenizer = cls(*init_inputs, **init_kwargs) File "/home/developer/.cache/huggingface/modules/transformers_modules/THUDM/chatglm2-6b/8fd7fba285f7171d3ae7ea3b35c53b6340501ed1/tokenization_chatglm.py", line 69, in init super().init(padding_side=padding_side, clean_up_tokenization_spaces=clean_up_tokenization_spaces, **kwargs) File "/home/developer/mambaforge/envs/CodeLlama/lib/python3.10/site-packages/transformers/tokenization_utils.py", line 366, in init self._add_tokens(self.all_special_tokens_extended, special_tokens=True) File "/home/developer/mambaforge/envs/CodeLlama/lib/python3.10/site-packages/transformers/tokenization_utils.py", line 454, in _add_tokens current_vocab = self.get_vocab().copy() File "/home/developer/.cache/huggingface/modules/transformers_modules/THUDM/chatglm2-6b/8fd7fba285f7171d3ae7ea3b35c53b6340501ed1/tokenization_chatglm.py", line 112, in get_vocab vocab = {self._convert_id_to_token(i): i for i in range(self.vocab_size)} File "/home/developer/.cache/huggingface/modules/transformers_modules/THUDM/chatglm2-6b/8fd7fba285f7171d3ae7ea3b35c53b6340501ed1/tokenization_chatglm.py", line 108, in vocab_size return self.tokenizer.n_words AttributeError: 'ChatGLMTokenizer' object has no attribute 'tokenizer'. Did you mean: 'tokenize'?

[phalexo]

Running your Jupyter notebook

File ~/.cache/huggingface/modules/transformers_modules/THUDM/chatglm2-6b/8fd7fba285f7171d3ae7ea3b35c53b6340501ed1/tokenization_chatglm.py:69, in ChatGLMTokenizer.init(self, vocab_file, padding_side, clean_up_tokenization_spaces, **kwargs) 68 def init(self, vocab_file, padding_side="left", clean_up_tokenization_spaces=False, **kwargs): ---> 69 super().init(padding_side=padding_side, clean_up_tokenization_spaces=clean_up_tokenization_spaces, **kwargs) 70 self.name = "GLMTokenizer" 72 self.vocab_file = vocab_file

File ~/mambaforge/envs/FinGPT/lib/python3.10/site-packages/transformers/tokenization_utils.py:366, in PreTrainedTokenizer.init(self, **kwargs) 362 self._added_tokens_encoder: Dict[str, int] = {k.content: v for v, k in self._added_tokens_decoder.items()} 364 # 4. If some of the special tokens are not part of the vocab, we add them, at the end. 365 # the order of addition is the same as self.SPECIAL_TOKENS_ATTRIBUTES following tokenizers --> 366 self._add_tokens(self.all_special_tokens_extended, special_tokens=True) 368 self._decode_use_source_tokenizer = False

File ~/mambaforge/envs/FinGPT/lib/python3.10/site-packages/transformers/tokenization_utils.py:454, in PreTrainedTokenizer._add_tokens(self, new_tokens, special_tokens) 452 if new_tokens is None: 453 return added_tokens --> 454 current_vocab = self.get_vocab().copy() 455 new_idx = len(current_vocab) # only call this once, len gives the last index + 1 456 for token in new_tokens:

File ~/.cache/huggingface/modules/transformers_modules/THUDM/chatglm2-6b/8fd7fba285f7171d3ae7ea3b35c53b6340501ed1/tokenization_chatglm.py:112, in ChatGLMTokenizer.get_vocab(self) 110 def get_vocab(self): 111 """ Returns vocab as a dict """ --> 112 vocab = {self._convert_id_to_token(i): i for i in range(self.vocab_size)} 113 vocab.update(self.added_tokens_encoder) 114 return vocab

File ~/.cache/huggingface/modules/transformers_modules/THUDM/chatglm2-6b/8fd7fba285f7171d3ae7ea3b35c53b6340501ed1/tokenization_chatglm.py:108, in ChatGLMTokenizer.vocab_size(self) 106 @property 107 def vocab_size(self): --> 108 return self.tokenizer.n_words

AttributeError: 'ChatGLMTokenizer' object has no attribute 'tokenizer'

[Weiyao-Li]

!pip install protobuf transformers==4.30.2 cpm_kernels torch>=2.0 gradio mdtex2html sentencepiece accelerate