This is the repository for AlignScore and its checkpoints, a metric for automatic factual consistency evaluation of text pairs. The metric is introduced in

AlignScore: Evaluating Factual Consistency with a Unified Alignment Function

Yuheng Zha, Yichi Yang, Ruichen Li and Zhiting Hu

ACL 2023

Code is at https://github.com/yuh-zha/AlignScore

What is factual consistency and its evaluation?

Facutual Consistency: For a given text pair (a, b), they are considered factual consistent if 1) all the information in b is also present in a; 2) b does not contradict a.
Evaluation: Show the degree of factual consistency between the context (text a) and the claim (text b).

Where is factual consistency evaluation applicable?

Summarization: document and summary
Paraphrase: sentence A and sentence B
Dialog: context and response
...

Leaderboard

We list the performance of AlignScore as well as other metrics here.

Rank	Metrics	SummaC*	TRUE**	Other-Spearman	Average	Paper	Code
1	AlignScore-large	88.6	83.8	49.3	73.9	-	-
2	AlignScore-base	87.4	82.5	44.9	71.6	-	-
3	QAFactEval	83.8	79.4	42.4	68.5	:page_facing_up:(Fabbri et al. 2022)	:octocat:
4	UniEval	84.6	78.0	41.5	68.0	:page_facing_up:(Zhong et al. 2022)	:octocat:
5	SummaC-CONV	81.0	78.7	34.2	64.6	:page_facing_up:(Laban et al. 2022)	:octocat:
6	BARTScore	80.9	73.4	34.8	63.0	:page_facing_up:(Yuan et al. 2022)	:octocat:
7	CTC	81.2	72.4	35.3	63.0	:page_facing_up:(Deng et al. 2022)	:octocat:
8	SummaC-ZS	79.0	78.2	30.4	62.5	:page_facing_up:(Laban et al. 2022)	:octocat:
9	ROUGE-2	78.1	72.4	27.9	59.5	:page_facing_up:(Lin 2004)	:octocat:
10	ROUGE-1	77.4	72.0	28.6	59.3	:page_facing_up:(Lin 2004)	:octocat:
11	ROUGE-L	77.3	71.8	28.3	59.1	:page_facing_up:(Lin 2004)	:octocat:
12	QuestEval	72.5	71.4	25.0	56.3	:page_facing_up:(Scialom et al. 2021)	:octocat:
13	BLEU	76.3	67.3	24.6	56.1	:page_facing_up:(Papineni et al. 2002)	:octocat:
14	DAE	66.8	65.7	35.1	55.8	:page_facing_up:(Goyal and Durrett 2020)	:octocat:
15	BLEURT	69.2	71.9	24.9	55.4	:page_facing_up:(Sellam et al. 2020)	:octocat:
16	BERTScore	72.1	68.6	21.9	54.2	:page_facing_up:(Zhang et al. 2020)	:octocat:
17	SimCSE	67.4	70.3	23.8	53.8	:page_facing_up:(Gao et al. 2021)	:octocat:
18	FactCC	68.8	62.7	21.2	50.9	:page_facing_up:(Kryscinski et al. 2020)	:octocat:
19	BLANC	65.1	64.0	14.4	47.8	:page_facing_up:(Vasilyev et al. 2020)	:octocat:
20	NER-Overlap	60.4	59.3	18.9	46.2	:page_facing_up:(Laban et al. 2022)	:octocat:
21	MNLI	47.9	60.4	3.1	37.2	:page_facing_up:(Williams et al. 2018)	:octocat:
22	FEQA	48.3	52.2	-1.9	32.9	:page_facing_up:(Durmus et al. 2020)	:octocat:

* SummaC: [Paper] | [Github]

** TRUE: [Paper] | [Github]

Installation

Our models are trained and evaluated using PyTorch 1.12.1. We recommend using this version to reproduce the results.

Please first install the right version of PyTorch before installing alignscore.
You can install alignscore by cloning this repository and pip install ..
After installing alignscore, please use python -m spacy download en_core_web_sm to install the required spaCy model (we use spaCy for sentenization).

Evaluating Factual Consistency

To evaluate the factual consistency of the claim w.r.t. the context, simply use the score method of AlignScore.

python

from alignscore import AlignScore

scorer = AlignScore(model='roberta-base', batch_size=32, device='cuda:0', ckpt_path='/path/to/checkpoint', evaluation_mode='nli_sp')
score = scorer.score(contexts=['hello world'], claims=['hello world'])

model: the backbone model of the metric. Now, we only provide the metric trained on RoBERTa

batch_size: the batch size of the inference

device: which device to run the metric

ckpt_path: the path to the checkpoint

evaluation_mode: choose from 'nli_sp', 'nli', 'bin_sp', 'bin'. nli and bin refer to the 3-way and binary classficiation head, respectively. sp indicates if the chunk-sentence splitting method is used. nli_sp is the default setting of AlignScore

Checkpoints

We provide two versions of the AlignScore checkpoints: AlignScore-base and AlignScore-large. The -base model is based on RoBERTa-base and has 125M parameters. The -large model is based on RoBERTa-large and has 355M parameters.

AlignScore-base: https://huggingface.co/yzha/AlignScore/resolve/main/AlignScore-base.ckpt

AlignScore-large: https://huggingface.co/yzha/AlignScore/resolve/main/AlignScore-large.ckpt

Training

You can use the above checkpoints directly for factual consistency evaluation. However, if you wish to train an alignment model from scratch / on your own data, use train.py.

python

python train.py --seed 2022 --batch-size 32 \
--num-epoch 3 --devices 0 1 2 3 \
--model-name roberta-large -- ckpt-save-path ./ckpt/ \
--data-path ./data/training_sets/ \
--max-samples-per-dataset 500000

--seed: the random seed for initialization

--batch-size: the batch size for training

--num-epoch: training epochs

--devices: which devices to train the metric, a list of GPU ids

--model-name: the backbone model name of the metric, default RoBERTa-large

--ckpt-save-path: the path to save the checkpoint

--training-datasets: the names of the training datasets

--data-path: the path to the training datasets

--max-samples-per-dataset: the maximum number of samples from a dataset

Benchmarking

Our benchmark includes the TRUE and SummaC benchmark as well as several popular factual consistency evaluation datasets.

To run the benchmark, a few additional dependencies are required and can be installed with pip install -r requirements.txt. Additionally, some depedencies are not available as packages and need to be downloaded manually (please see python benchmark.py --help for instructions).

Note installing summac may cause dependency conflicts with alignscore. Please reinstall alignscore to force the correct dependency versions.

The relevant arguments for evaluating AlignScore are:

--alignscore: evaluation the AlignScore metric

--alignscore-model: the name of the backbone model (either 'roberta-base' or 'roberta-large')

--alignscore-ckpt: the path to the saved checkpoint

--alignscore-eval-mode: the evaluation mode, defaults to nli_sp

--device: which device to run the metric, defaults to cuda:0

--tasks: which tasks to benchmark, e.g., SummEval, QAGS-CNNDM, ...

For the baselines, please see python benchmark.py --help for details.