Paitesanshi/LLM_MBTI

1. Install

First, install all packages with:

pip install -r requirements.txt

2. Get MBTI for Single Character

run get_llms_mbti.py to download models and test target character's mbti.

You can specify models by following code:

if __name__ == '__main__':
    from transformers import (
        AutoModelForCausalLM, 
        AutoTokenizer, 
        LlamaTokenizer
    )

    models = [
        'baichuan-inc/Baichuan-7B',
        'bigscience/bloom-7b1',
    ]

    tokenizers = [
        'baichuan-inc/Baichuan-7B',
        'bigscience/bloom-7b1',
    ]
    character="Beethoven"

    ...

3. Get MBTI for 20 Famous Characters

run eval_roles_mbti_zh.py to test 20 famous characters' mbti with Chinese prompts. run eval_roles_mbti_en.py to test 20 famous characters' mbti with English prompts.

The results are saved in pop20_mbti_parse_zh.json and pop20_mbti_parse_en.json respectively. The overall results are summarized in Table pop20_results_parse_mbti_zh.csv and pop20_results_parse_acc_zh.csv. The same applies to the English version.

4. Get MBTI for 16 MBTI Types

run eval_general_mbti_zh.py to test 16 MBTI types with Chinese prompts. run eval_general_mbti_en.py to test 16 MBTI types with English prompts.

The results are saved in 16_mbti_zh.json and 16_mbti_en.json respectively. The overall results are summarized in Table 16_results_mbti_zh.csv and 16_results_mbti_acc_zh.csv. The same applies to the English version.