MiniCheck-DeBERTa-v3-Large

This is a fact-checking model from our work:

📃 MiniCheck: Efficient Fact-Checking of LLMs on Grounding Documents (EMNLP 2024, GitHub Repo)

The model is based on DeBERTa-v3-Large that predicts a binary label - 1 for supported and 0 for unsupported. The model is doing predictions on the sentence-level. It takes as input a document and a sentence and determine whether the sentence is supported by the document: MiniCheck-Model(document, claim) -> {0, 1}

MiniCheck-DeBERTa-v3-Large is fine tuned from microsoft/deberta-v3-large (He et al., 2023) on the combination of 35K data:

• 21K ANLI data (Nie et al., 2020)
• 14K synthetic data generated from scratch in a structed way (more details in the paper).

Model Variants

• bespokelabs/Bespoke-Minicheck-7B (Model Size: 7B)
• lytang/MiniCheck-Flan-T5-Large (Model Size: 0.8B)
• lytang/MiniCheck-RoBERTa-Large (Model Size: 0.4B)

Model Performance

The performance of these models is evaluated on our new collected benchmark (unseen by our models during training), LLM-AggreFact, from 11 recent human annotated datasets on fact-checking and grounding LLM generations. MiniCheck-DeBERTa-v3-Large outperform all exisiting specialized fact-checkers with a similar scale. See full results in our work.

Note: We only evaluated the performance of our models on real claims -- without any human intervention in any format, such as injecting certain error types into model-generated claims. Those edited claims do not reflect LLMs' actual behaviors.

Model Usage Demo

Please run the following command to install the MiniCheck package and all necessary dependencies.

pip install "minicheck @ git+https://github.com/Liyan06/MiniCheck.git@main"

Below is a simple use case

from minicheck.minicheck import MiniCheck
import os
os.environ["CUDA_VISIBLE_DEVICES"] = "0"

doc = "A group of students gather in the school library to study for their upcoming final exams."
claim_1 = "The students are preparing for an examination."
claim_2 = "The students are on vacation."

# model_name can be one of ['roberta-large', 'deberta-v3-large', 'flan-t5-large', 'Bespoke-MiniCheck-7B']
scorer = MiniCheck(model_name='deberta-v3-large', cache_dir='./ckpts')
pred_label, raw_prob, _, _ = scorer.score(docs=[doc, doc], claims=[claim_1, claim_2])

print(pred_label) # [1, 0]
print(raw_prob)   # [0.9786180257797241, 0.01138285268098116]

Test on our LLM-AggreFact Benchmark

import pandas as pd
from datasets import load_dataset
from minicheck.minicheck import MiniCheck
import os
os.environ["CUDA_VISIBLE_DEVICES"] = "0"

# load 29K test data
df = pd.DataFrame(load_dataset("lytang/LLM-AggreFact")['test'])
docs = df.doc.values
claims = df.claim.values

scorer = MiniCheck(model_name='deberta-v3-large', cache_dir='./ckpts')
pred_label, raw_prob, _, _ = scorer.score(docs=docs, claims=claims)  # ~ 800 docs/min, depending on hardware

To evalaute the result on the benchmark

from sklearn.metrics import balanced_accuracy_score

df['preds'] = pred_label
result_df = pd.DataFrame(columns=['Dataset', 'BAcc'])
for dataset in df.dataset.unique():
    sub_df = df[df.dataset == dataset]
    bacc = balanced_accuracy_score(sub_df.label, sub_df.preds) * 100
    result_df.loc[len(result_df)] = [dataset, bacc]
result_df.loc[len(result_df)] = ['Average', result_df.BAcc.mean()]
result_df.round(1)

Citation

@InProceedings{tang-etal-2024-minicheck,
  title = {MiniCheck: Efficient Fact-Checking of LLMs on Grounding Documents},
  author = {Liyan Tang and Philippe Laban and Greg Durrett},
  booktitle = {Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing},
  year = {2024},
  publisher = {Association for Computational Linguistics},
  url = {https://arxiv.org/pdf/2404.10774}
}