README.md

---
language:
- en
- ar
- zh
- nl
- fr
- de
- hi
- in
- it
- ja
- pt
- ru
- es
- vi
- multilingual
license: apache-2.0
datasets:
- unicamp-dl/mmarco
widget:
- text: Python ist eine universelle, �blicherweise interpretierte, h�here Programmiersprache.
    Sie hat den Anspruch, einen gut lesbaren, knappen Programmierstil zu f�rdern.
    So werden beispielsweise Bl�cke nicht durch geschweifte Klammern, sondern durch
    Einr�ckungen strukturiert.
---

# doc2query/msmarco-14langs-mt5-base-v1

This is a [doc2query](https://arxiv.org/abs/1904.08375) model based on mT5 (also known as [docT5query](https://cs.uwaterloo.ca/~jimmylin/publications/Nogueira_Lin_2019_docTTTTTquery-v2.pdf)). It was trained on all 14 languages of [mMARCO dataset](https://github.com/unicamp-dl/mMARCO), i.e. you can input a passage in any of the 14 languages, and it will generate a query in the same language.

It can be used for:
- **Document expansion**: You generate for your paragraphs 20-40 queries and index the paragraphs and the generates queries in a standard BM25 index like Elasticsearch, OpenSearch, or Lucene. The generated queries help to close the lexical gap of lexical search, as the generate queries contain synonyms. Further, it re-weights words giving important words a higher weight even if they appear seldomn in a paragraph. In our [BEIR](https://arxiv.org/abs/2104.08663) paper we showed that BM25+docT5query is a powerful search engine. In the [BEIR repository](https://github.com/beir-cellar/beir) we have an example how to use docT5query with Pyserini.
- **Domain Specific Training Data Generation**: It can be used to generate training data to learn an embedding model. In our [GPL-Paper](https://arxiv.org/abs/2112.07577) / [GPL Example on SBERT.net](https://www.sbert.net/examples/domain_adaptation/README.html#gpl-generative-pseudo-labeling) we have an example how to use the model to generate (query, text) pairs for a given collection of unlabeled texts. These pairs can then be used to train powerful dense embedding models.

## Usage
```python
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
import torch

model_name = 'doc2query/msmarco-14langs-mt5-base-v1'
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSeq2SeqLM.from_pretrained(model_name)

text = "Python ist eine universelle, �blicherweise interpretierte, h�here Programmiersprache. Sie hat den Anspruch, einen gut lesbaren, knappen Programmierstil zu f�rdern. So werden beispielsweise Bl�cke nicht durch geschweifte Klammern, sondern durch Einr�ckungen strukturiert."


def create_queries(para):
    input_ids = tokenizer.encode(para, return_tensors='pt')
    with torch.no_grad():
        # Here we use top_k / top_k random sampling. It generates more diverse queries, but of lower quality
        sampling_outputs = model.generate(
            input_ids=input_ids,
            max_length=64,
            do_sample=True,
            top_p=0.95,
            top_k=10, 
            num_return_sequences=5
            )
        
        # Here we use Beam-search. It generates better quality queries, but with less diversity
        beam_outputs = model.generate(
            input_ids=input_ids, 
            max_length=64, 
            num_beams=5, 
            no_repeat_ngram_size=2, 
            num_return_sequences=5, 
            early_stopping=True
        )


    print("Paragraph:")
    print(para)
    
    print("\nBeam Outputs:")
    for i in range(len(beam_outputs)):
        query = tokenizer.decode(beam_outputs[i], skip_special_tokens=True)
        print(f'{i + 1}: {query}')

    print("\nSampling Outputs:")
    for i in range(len(sampling_outputs)):
        query = tokenizer.decode(sampling_outputs[i], skip_special_tokens=True)
        print(f'{i + 1}: {query}')

create_queries(text)

```

**Note:** `model.generate()` is non-deterministic for top_k/top_n sampling. It produces different queries each time you run it.

## Training
This model fine-tuned [google/mt5-base](https://huggingface.co/google/mt5-base) for 525k training steps on all 14 languages from [mMARCO dataset](https://github.com/unicamp-dl/mMARCO). For the  training script, see the `train_script.py` in this repository.

The input-text was truncated to 320 word pieces. Output text was generated up to 64 word pieces. 

This model was trained on a (query, passage) from the [mMARCO dataset](https://github.com/unicamp-dl/mMARCO).