law-llm

/

law-glm-10b

like 3

Hi,

Thank you for reporting the issue. I will look into it in detail later as I am currently occupied with other tasks.
In the meantime, I believe you can find the file here: https://huggingface.co/THUDM/glm-10b-chinese/blob/main/tokenization_glm.py.

Would that help?

yeah,i try it，but this https://huggingface.co/THUDM/glm-10b-chinese/blob/main/tokenization_glm.py. website cannot be accessed ,it's 404

I just tried it and could open it. Did you use a VPN?

thanks for your reply！yeah,i used VPN but I can access it now. i want to download this model files completely, however i fear it cannot run in my local machine(windows) , so how much GPU memory can deal it, my machine 's GPU is 4060, can run it successful?

Hi,

Running the model for inference might be challenging for a 4060 with 8GB of vRAM. You might want to try using int8/int4 mode for inference. We previously used multiple A100-40GB GPUs for training and inference.