---
license: apache-2.0
---

# Spanish Bert2Bert fine-tuned on Quora question pairs dataset

Fine-tuning of a [question generator model](https://huggingface.co/mrm8488/bert2bert-spanish-question-generation) into a paraphraser model using a poor-man's translation of the Quora question pairs dataset. It basically rephrases questions into similar questions. Non interrogative sentences are not handled very well.

- Original models: [mrm8488/bert2bert-spanish-question-generation](https://huggingface.co/mrm8488/bert2bert-spanish-question-generation?text=Manuel+vive+en+Murcia%2C+Espa%C3%B1a), which is based on [dccuchile/bert-base-spanish-wwm-cased](https://huggingface.co/dccuchile/bert-base-spanish-wwm-cased) (?).
- Custom database: "Poor-man's"  translation of duplicated questions in Quora (translated with [Helsinki-NLP/opus-mt-en-es](https://huggingface.co/Helsinki-NLP/opus-mt-en-es))