We cast this task as a pairwise classification task, where the objective is to compare two versions of the same claim and determine which one is better.

Usage

python

from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch

tokenizer = AutoTokenizer.from_pretrained("gabski/bert-relative-claim-quality")
model = AutoModelForSequenceClassification.from_pretrained("gabski/bert-relative-claim-quality")
claim_1 = 'Smoking marijuana is less harmfull then smoking cigarettes.'
claim_2 = 'Smoking marijuana is less harmful than smoking cigarettes.'
model_input = tokenizer(claim_1,claim_2, return_tensors='pt')
model_outputs = model(**model_input)

outputs = torch.nn.functional.softmax(model_outputs.logits, dim = -1)
print(outputs)