Metadata-Version: 2.4
Name: minkarr
Version: 0.1.7
Summary: A minimal implementation of KaRR knowledge assessment method for Large Language Models (LLMs)
Author-email: Hichem Ammar Khodja <hichem.ammarkhodja@orange.com>
Requires-Python: >=3.8
Requires-Dist: jsonlines>=4.0.0
Requires-Dist: nltk>=3.9.1
Requires-Dist: setuptools>=70.0.0
Requires-Dist: transformers>=4.46.0
Requires-Dist: wikidataintegrator>=0.9.30
Description-Content-Type: text/markdown

# Statistical Knowledge Assessment for Large Language Models
A minimal implementation of KaRR knowledge assessment method from the following paper:

> [**Statistical Knowledge Assessment for Large Language Models**](https://arxiv.org/abs/2305.10519),            
> Qingxiu Dong, Jingjing Xu, Lingpeng Kong, Zhifang Sui, Lei Li   
> *arXiv preprint ([arxiv_version](https://arxiv.org/abs/2305.10519))*   

This is a fork of the [official implementation](https://github.com/dqxiu/KAssess) released by the authors.

## How to use?

First, create a new virtual environment, then install [Pytorch-CUDA](https://pytorch.org/get-started/locally/), and finally install **minkarr** using the command:

```bash
pip install minkarr
```

Here is a simple example of **how to quantify the knowledge of a fact by an LLM using KaRR**
```python
from minkarr import KaRR
from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "gpt2"
device = "cuda"
model = AutoModelForCausalLM.from_pretrained(model_name).cuda()
tokenizer = AutoTokenizer.from_pretrained(model_name)

karr = KaRR(model, tokenizer, device)

# Testing the fact: (France, capital, Paris)
# You can find other facts by looking into Wikidata
fact = ("Q142", "P36", "Q90")

karr, does_know = karr.compute(fact)
print("Fact %s" % str(fact))
print("KaRR = %s" % karr)
ans = "Yes" if does_know else "No"
print("According to KaRR, does the model knows this fact? Answer: %s" % ans)
# Output:
# KaRR = 3.338972442145268
# According to KaRR, does the model knows this fact? Answer: No
```

## Difference with original repo

- Easy-to-use
- Cleaner code
- Minimalistic implementation: I kept only the portion of the code needed to compute KaRR and removed the rest
- This implementation can compute KaRR on a single fact (the original implementation went through all facts)

## Change storage location

Set the environment variable `STORAGE_FOLDER` to choose where to store the data that MinKaRR downloads.

## Citation
Cite the original authors using:
```
@misc{dong2023statistical,
      title={Statistical Knowledge Assessment for Large Language Models}, 
      author={Qingxiu Dong and Jingjing Xu and Lingpeng Kong and Zhifang Sui and Lei Li},
      year={2023},
      journal = {Proceedings of NeurIPS},
}
```



