google

/

gemma-2b

like 679

same error for me as well

ValueError Traceback (most recent call last)
in <cell line: 4>()
2 from transformers import AutoTokenizer, AutoModelForCausalLM
3
----> 4 tokenizer = AutoTokenizer.from_pretrained("google/gemma-2b")
5 model = AutoModelForCausalLM.from_pretrained("google/gemma-2b", device_map="auto")
6

/usr/local/lib/python3.10/dist-packages/transformers/models/auto/tokenization_auto.py in from_pretrained(cls, pretrained_model_name_or_path, *inputs, **kwargs)
809 tokenizer_class.register_for_auto_class()
810 return tokenizer_class.from_pretrained(
--> 811 pretrained_model_name_or_path, *inputs, trust_remote_code=trust_remote_code, **kwargs
812 )
813 elif config_tokenizer_class is not None:

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

reinstall transformers

pip uninstall transformers

pip install transformers

pip install -U transformers should solve the issue. If not, make sure to create a fresh new env

tokenizer = AutoTokenizer.from_pretrained("google/gemma-2b", use_auth_token=True)
model = AutoModelForCausalLM.from_pretrained("google/gemma-2b", device_map="auto")

''''''''
tokenizer not imported in this and below error
already updated tokenizer but got same error

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

Hi there! Did you upgrade transformers by doing pip install -U transformer?

did but got the same error as below:

ValueError Traceback (most recent call last)
in <cell line: 5>()
3
4 # tokenizer = GemmaTokenizer.from_preset("google/gemma-2b", use_auth_token=True)
----> 5 tokenizer = AutoTokenizer.from_pretrained("google/gemma-7b", use_auth_token=True)
6 model = AutoModelForCausalLM.from_pretrained("google/gemma-2b", device_map="auto")

/usr/local/lib/python3.10/dist-packages/transformers/models/auto/tokenization_auto.py in from_pretrained(cls, pretrained_model_name_or_path, *inputs, **kwargs)
697 Will be passed to the Tokenizer __init__() method. Can be used to set special tokens like
698 bos_token, eos_token, unk_token, sep_token, pad_token, cls_token, mask_token,
--> 699 additional_special_tokens. See parameters in the __init__() for more details.
700
701 Examples:

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

I am running in google colab, which has python 3.10

issue resolved, after updating python from 3.10 to 3.11

Hi, just restart the session again but before that,
remove any pip install transformers command and replace it with below line.
!pip -q install git+https://github.com/huggingface/transformers.git

Install other packages like trl or peft seperately using !pip install trl peft
It started working for me..

it worked for me too! thanks

it worked for me too! thanks

Hey! Can you please share your colab code? I'm getting multiple errors, it will be helpful thanks

reinstall transformers

pip uninstall transformers

pip install transformers

After reinstallation, the versation of transformers still keep same

transformers-4.30.2
# python
Python 3.7.3

@WQW You need to upgrade it

pip install --upgrade transformers

A simple install will used the cached version of your already installed package

Colab comes with an older transformers version pre-installed. If you install/upgrade the last version, you will need to restart the runtime so it picks the newly installed version

@WQW You need to upgrade it

pip install --upgrade transformers

A simple install will used the cached version of your already installed package

I've already updated to 4.38

pip install transformers==4.38.2 and it works