Models: bag-of-words approaches with WordPiece tokenization provided by HuggingFace Tokenizer Integration
This page documents regression experiments on the MS MARCO document ranking task, which is integrated into Anserini's regression testing framework.
Here we are using WordPiece tokenization (i.e., from BERT) with the following tokenizer from HuggingFace bert-base-uncased
: .
In general, effectiveness is lower than with "standard" Lucene tokenization for two reasons: (1) we're losing stemming, and (2) some terms are chopped into less meaningful subwords.
The exact configurations for these regressions are stored in this YAML file. Note that this page is automatically generated from this template as part of Anserini's regression pipeline, so do not modify this page directly; modify the template instead.
From one of our Waterloo servers (e.g., orca
), the following command will perform the complete regression, end to end:
python src/main/python/run_regression.py --index --verify --search --regression msmarco-v1-doc.wp-hgf
Typical indexing command:
bin/run.sh io.anserini.index.IndexCollection \
-collection JsonCollection \
-input /path/to/msmarco-doc \
-generator DefaultLuceneDocumentGenerator \
-index indexes/lucene-inverted.msmarco-v1-doc.wp-hgf/ \
-threads 7 -storePositions -storeDocvectors -storeRaw -analyzeWithHuggingFaceTokenizer bert-base-uncased \
>& logs/log.msmarco-doc &
The directory /path/to/msmarco-passage-wp/
should be a directory containing the corpus in Anserini's jsonl format.
For additional details, see explanation of common indexing options.
Topics and qrels are stored here, which is linked to the Anserini repo as a submodule. The regression experiments here evaluate on the 5193 dev set questions.
After indexing has completed, you should be able to perform retrieval as follows:
bin/run.sh io.anserini.search.SearchCollection \
-index indexes/lucene-inverted.msmarco-v1-doc.wp-hgf/ \
-topics tools/topics-and-qrels/topics.msmarco-doc.dev.txt \
-topicReader TsvInt \
-output runs/run.msmarco-doc.bm25-default.topics.msmarco-doc.dev.txt \
-bm25 -analyzeWithHuggingFaceTokenizer bert-base-uncased &
Evaluation can be performed using trec_eval
:
bin/trec_eval -c -m map tools/topics-and-qrels/qrels.msmarco-doc.dev.txt runs/run.msmarco-doc.bm25-default.topics.msmarco-doc.dev.txt
bin/trec_eval -c -M 100 -m recip_rank tools/topics-and-qrels/qrels.msmarco-doc.dev.txt runs/run.msmarco-doc.bm25-default.topics.msmarco-doc.dev.txt
bin/trec_eval -c -m recall.100 tools/topics-and-qrels/qrels.msmarco-doc.dev.txt runs/run.msmarco-doc.bm25-default.topics.msmarco-doc.dev.txt
bin/trec_eval -c -m recall.1000 tools/topics-and-qrels/qrels.msmarco-doc.dev.txt runs/run.msmarco-doc.bm25-default.topics.msmarco-doc.dev.txt
With the above commands, you should be able to reproduce the following results:
AP@1000 | BM25 (default) |
---|---|
MS MARCO Doc: Dev | 0.2234 |
RR@100 | BM25 (default) |
MS MARCO Doc: Dev | 0.2227 |
R@100 | BM25 (default) |
MS MARCO Doc: Dev | 0.7031 |
R@1000 | BM25 (default) |
MS MARCO Doc: Dev | 0.8743 |