bbm25-haystack 0.1.1

Haystack 2.x In-memory Document Store with Enhanced Efficiency

Better BM25 In-Memory Document Store

An in-memory document store is a great starting point for prototyping and debugging before migrating to production-grade stores like Elasticsearch. However, the original implementation of BM25 retrieval recreates an inverse index for the entire document store on every new search. Furthermore, the tokenization method is primitive, only permitting splitters based on regular expressions, making localization and domain adaptation challenging. Therefore, this implementation is a slight upgrade to the default BM25 in-memory document store by implementing incremental index update and incorporation of SentencePiece statistical sub-word tokenization.

Installation

$ pip install bbm25-haystack

Alternatively, you can clone the repository and build from source to be able to reflect changes to the source code:

$ git clone https://github.com/Guest400123064/bbm25-haystack.git
$ cd bbm25-haystack
$ pip install -e .

Usage

The initializer takes three BM25+ hyperparameters, namely k1, b, and delta, one path to a trained SentencePiece tokenizer .model file, and a filtering logic flag (see below). All parameters are optional. The default tokenizer is directly copied from LLaMA-2-7B-32K tokenizer with a vocab size of 32,000.

from haystack import Document
from bbm25_haystack import BetterBM25DocumentStore, BetterBM25Retriever


document_store = BetterBM25DocumentStore()
document_store.write_documents([
   Document(content="There are over 7,000 languages spoken around the world today."),
   Document(content="Elephants have been observed to behave in a way that indicates a high level of self-awareness, such as recognizing themselves in mirrors."),
   Document(content="In certain parts of the world, like the Maldives, Puerto Rico, and San Diego, you can witness the phenomenon of bioluminescent waves.")
])

retriever = BetterBM25Retriever(document_store)
retriever.run(query="How many languages are spoken around the world today?")

Filtering Logic

The current document store uses document_matches_filter shipped with Haystack to perform filtering by default, which is the same as InMemoryDocumentStore except that it is DOES NOT support legacy operator names.

However, there is also an alternative filtering logic shipped with this implementation that is more conservative (and unstable at this point). To use this alternative logic, initialize the document store with haystack_filter_logic=False Please find comments and implementation details in filters.py. TL;DR:

• Comparison with None, i.e., missing values, involved will always return False, no matter the document attribute value or filter value.
• Comparison with DataFrame is always prohibited to reduce surprises.
• No implicit datetime conversion from string values.

These differences lead to a few caveats. Firstly, some test cases are overridden to take into account the different expectations. However, this means that passed, non-overridden tests may not be faithful, i.e., the filters behave in the same way as the old implementation while different behaviors are expected. Further, the negation logic needs to be considered again because False can now issue from both input check and the actual comparisons. But I think having input processing and comparisons separated makes the filtering behavior more transparent.

Search Quality Evaluation

This repo has a simple script to help evaluate the search quality over BEIR benchmark. You need to clone the repository (you can also manually download the script and place it under a folder named scripts) and you have to install additional dependencies to run the script.

$ pip install beir

To run the script, you may want to specify the dataset name and BM25 hyperparameters. For example:

$ python scripts/benchmark_beir.py --datasets scifact arguana --bm25-k1 1.2 --output eval.csv

It automatically downloads the benchmarking dataset to benchmarks/beir, where benchmarks is at the same level as scripts. You may also check the help page for more information.

$ python scripts/benchmark_beir.py --help

New benchmarking scripts are expected to be added in the future.

License

bbm25-haystack is distributed under the terms of the Apache-2.0 license.