haystack/tutorials/Tutorial6_Better_Retrieval_via_DPR.py
Sara Zan 91cafb49bb
Improve tutorials' output (#1694)
* Modify __str__ and __repr__ for Document and Answer

* Rename QueryClassifier in Tutorial11

* Improve the output of tutorial1

* Make the output of Tutorial8 a bit less dense

* Add a print_questions util to print the output of question generating pipelines

* Replace custom printing with the new utility in Tutorial13

* Ensure all output is printed with minimal details in Tutorial14 and add some titles

* Minor change to print_answers

* Make tutorial3's output the same as tutorial1

* Add __repr__ to Answer and fix to_dict()

* Fix a bug in the Document and Answer's __str__ method

* Improve print_answers, print_documents and print_questions

* Using print_answers in Tutorial7 and fixing typo in the utils

* Remove duplicate line in Tutorial12

* Use print_answers in Tutorial4

* Add explanation of what the documents in the output of the basic QA pipeline are

* Move the fields constant into print_answers

* Normalize all 'minimal' to 'minimum' (they were mixed up)

* Improve the sample output to include all fields from Document and Answer

Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
2021-11-09 15:09:26 +01:00

78 lines
4.2 KiB
Python
Executable File

from haystack.document_stores import FAISSDocumentStore, MilvusDocumentStore
from haystack.utils import clean_wiki_text, print_answers, launch_milvus, convert_files_to_dicts, fetch_archive_from_http
from haystack.nodes import FARMReader, DensePassageRetriever
def tutorial6_better_retrieval_via_dpr():
# OPTION 1: FAISS is a library for efficient similarity search on a cluster of dense vectors.
# The FAISSDocumentStore uses a SQL(SQLite in-memory be default) document store 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
document_store = FAISSDocumentStore(faiss_index_factory_str="Flat")
# OPTION2: Milvus is an open source database library that is also optimized for vector similarity searches like FAISS.
# Like FAISS it has both a "Flat" and "HNSW" mode but it outperforms FAISS when it comes to dynamic data management.
# It does require a little more setup, however, as it is run through Docker and requires the setup of some config files.
# See https://milvus.io/docs/v1.0.0/milvus_docker-cpu.md
# launch_milvus()
# document_store = MilvusDocumentStore()
# ## Preprocessing of documents
# Let's first get some documents that we want to query
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 containing documents that can be indexed to our datastore
dicts = convert_files_to_dicts(dir_path=doc_dir, clean_func=clean_wiki_text, split_paragraphs=True)
# Now, let's write the docs to our DB.
document_store.write_documents(dicts)
### Retriever
retriever = DensePassageRetriever(document_store=document_store,
query_embedding_model="facebook/dpr-question_encoder-single-nq-base",
passage_embedding_model="facebook/dpr-ctx_encoder-single-nq-base",
max_seq_len_query=64,
max_seq_len_passage=256,
batch_size=2,
use_gpu=True,
embed_title=True,
use_fast_tokenizers=True
)
# Important:
# Now that after we have the DPR initialized, we need to call update_embeddings() to iterate over all
# previously indexed documents and update their embedding representation.
# While this can be a time consuming operation (depending on corpus size), it only needs to be done once.
# At query time, we only need to embed the query and compare it the existing doc embeddings which is very fast.
document_store.update_embeddings(retriever)
### Reader
# Load a local model or any of the QA models on
# Hugging Face's model hub (https://huggingface.co/models)
reader = FARMReader(model_name_or_path="deepset/roberta-base-squad2", use_gpu=True)
### Pipeline
from haystack.pipelines import ExtractiveQAPipeline
pipe = ExtractiveQAPipeline(reader, retriever)
## Voilà! Ask a question!
prediction = pipe.run(
query="Who is the father of Arya Stark?", params={"Retriever": {"top_k": 10}, "Reader": {"top_k": 5}}
)
# prediction = pipe.run(query="Who created the Dothraki vocabulary?", params={"Reader": {"top_k": 5}})
# prediction = pipe.run(query="Who is the sister of Sansa?", params={"Reader": {"top_k": 5}})
print_answers(prediction, details="minimum")
if __name__ == "__main__":
tutorial6_better_retrieval_via_dpr()
# This Haystack script was made with love by deepset in Berlin, Germany
# Haystack: https://github.com/deepset-ai/haystack
# deepset: https://deepset.ai/