| This is a version of paraphrase detector by DeepPavlov ([details in the documentation](http://docs.deeppavlov.ai/en/master/features/overview.html#ranking-model-docs)) ported to the `Transformers` format. | |
| All credit goes to the authors of DeepPavlov. | |
| The model has been trained on the dataset from http://paraphraser.ru/. | |
| It classifies texts as paraphrases (class 1) or non-paraphrases (class 0). | |
| ```python | |
| import torch | |
| from transformers import AutoModelForSequenceClassification, BertTokenizer | |
| model_name = 'cointegrated/rubert-base-cased-dp-paraphrase-detection' | |
| model = AutoModelForSequenceClassification.from_pretrained(model_name).cuda() | |
| tokenizer = BertTokenizer.from_pretrained(model_name) | |
| text1 = 'Сегодня на улице хорошая погода' | |
| text2 = 'Сегодня на улице отвратительная погода' | |
| batch = tokenizer(text1, text2, return_tensors='pt').to(model.device) | |
| with torch.inference_mode(): | |
| proba = torch.softmax(model(**batch).logits, -1).cpu().numpy() | |
| print(proba) | |
| # [[0.44876656 0.5512334 ]] | |
| ``` | |