Haystack docs home page

Long-Form Question Answering

Open In Colab

Prepare environment

Colab: Enable the GPU runtime

Make sure you enable the GPU runtime to experience decent speed in this tutorial.
Runtime -> Change Runtime type -> Hardware accelerator -> GPU

# Make sure you have a GPU running
# Install the latest master of Haystack
!pip install git+https://github.com/deepset-ai/haystack.git
from haystack.preprocessor.cleaning import clean_wiki_text
from haystack.preprocessor.utils import convert_files_to_dicts, fetch_archive_from_http
from haystack.generator.transformers import Seq2SeqGenerator

Document Store

FAISS is a library for efficient similarity search on a cluster of dense vectors. The FAISSDocumentStore uses a SQL(SQLite in-memory be default) database under-the-hood to store the document text and other meta data. The vector embeddings of the text are indexed on a FAISS Index that later is queried for searching answers. The default flavour of FAISSDocumentStore is "Flat" but can also be set to "HNSW" for faster search at the expense of some accuracy. Just set the faiss_index_factor_str argument in the constructor. For more info on which suits your use case: https://github.com/facebookresearch/faiss/wiki/Guidelines-to-choose-an-index

from haystack.document_store.faiss import FAISSDocumentStore

document_store = FAISSDocumentStore(vector_dim=128, faiss_index_factory_str="Flat")

Cleaning & indexing documents

Similarly to the previous tutorials, we download, convert and index some Game of Thrones articles to our DocumentStore

# Let's first get some files that we want to use
doc_dir = "data/article_txt_got"
s3_url = "https://s3.eu-central-1.amazonaws.com/deepset.ai-farm-qa/datasets/documents/wiki_gameofthrones_txt.zip"
fetch_archive_from_http(url=s3_url, output_dir=doc_dir)

# Convert files to dicts
dicts = convert_files_to_dicts(dir_path=doc_dir, clean_func=clean_wiki_text, split_paragraphs=True)
# Now, let's write the dicts containing documents to our DB.

Initialize Retriever and Reader/Generator


Here: We use a RetribertRetriever and we invoke update_embeddings to index the embeddings of documents in the FAISSDocumentStore

from haystack.retriever.dense import EmbeddingRetriever

retriever = EmbeddingRetriever(document_store=document_store,


Before we blindly use the RetribertRetriever let's empirically test it to make sure a simple search indeed finds the relevant documents.

from haystack.utils import print_answers, print_documents
from haystack.pipeline import DocumentSearchPipeline

p_retrieval = DocumentSearchPipeline(retriever)
res = p_retrieval.run(
    query="Tell me something about Arya Stark?",
print_documents(res, max_text_len=512)


Similar to previous Tutorials we now initalize our reader/generator.

Here we use a Seq2SeqGenerator with the yjernite/bart_eli5 model (see: https://huggingface.co/yjernite/bart_eli5)

generator = Seq2SeqGenerator(model_name_or_path="yjernite/bart_eli5")


With a Haystack Pipeline you can stick together your building blocks to a search pipeline. Under the hood, Pipelines are Directed Acyclic Graphs (DAGs) that you can easily customize for your own use cases. To speed things up, Haystack also comes with a few predefined Pipelines. One of them is the GenerativeQAPipeline that combines a retriever and a reader/generator to answer our questions. You can learn more about Pipelines in the docs.

from haystack.pipeline import GenerativeQAPipeline
pipe = GenerativeQAPipeline(generator, retriever)

Voilà! Ask a question!

pipe.run(query="Why did Arya Stark's character get portrayed in a television adaptation?", top_k_retriever=1)
pipe.run(query="What kind of character does Arya Stark play?", top_k_retriever=1)

About us

This Haystack notebook was made with love by deepset in Berlin, Germany

We bring NLP to the industry via open source! Our focus: Industry specific language models & large scale QA systems.

Some of our other work:

Get in touch: Twitter | LinkedIn | Slack | GitHub Discussions | Website

By the way: we're hiring!