questions on mistral 7B with CoT

[twni2016]

Hi authors,

Thank you for your nice library! I am trying to use your library to run mistral 7B with CoT on gsm8k. I have several questions on the code when using HFModel:

• Which mistral 7B model was used in your paper?
• I tried mistralai/Mistral-7B-v0.3 and found the eos_token_id is no longer 13 Based on this, I think using eos_token_id = ["\n\n", ".\n", "\n", ".\n\n"] like Llama3Model is better?
• Is this a typo that max_batch_size=batch_sizeis not added when calling HFModel?

Thank you for your time!

[Ber666]

Hi! Thanks for your questions.

It was v0.1. See the command line here. Sorry that the library was refactored and this information was lost.

We will fix other issues you mentioned.