Skip to content

Latest commit

 

History

History
103 lines (72 loc) · 7.02 KB

regressions-beir-v1.0.0-cqadupstack-webmasters-splade-distil-cocodenser-medium.md

File metadata and controls

103 lines (72 loc) · 7.02 KB
Raw

Anserini Regressions: BEIR (v1.0.0) — CQADupStack-webmasters

Model: SPLADE-distil CoCodenser Medium

This page describes regression experiments, integrated into Anserini's regression testing framework, using SPLADE-distil CoCodenser Medium on BEIR (v1.0.0) — CQADupStack-webmasters. SPLADE-distil CoCodenser Medium is an intermediate model version between SPLADEv2 and SPLADE++, where the model used distillation (as in SPLADEv2), but started with the CoCondenser pre-trained model. See the official SPLADE repo for more details; the model itself can be download here.

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 and then run bin/build.sh to rebuild the documentation.

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 beir-v1.0.0-cqadupstack-webmasters-splade-distil-cocodenser-medium

Corpus

We make available a version of the BEIR-v1.0.0 cqadupstack-webmasters corpus that has already been processed with SPLADE-distil CoCodenser Medium, i.e., gone through document expansion and term reweighting. Thus, no neural inference is involved. For details on how to train SPLADE-distil CoCodenser Medium and perform inference, please see guide provided by Naver Labs Europe.

Download the corpus and unpack into collections/:

wget https://rgw.cs.uwaterloo.ca/JIMMYLIN-bucket0/data/beir-v1.0.0-splade_distil_cocodenser_medium-cqadupstack-webmasters.tar -P collections/
tar xvf collections/beir-v1.0.0-splade_distil_cocodenser_medium-cqadupstack-webmasters.tar -C collections/

To confirm, the tarball is 8.9 MB and has MD5 checksum 9c5a181e03cbc7f13abd0e0e4bf9158e.

With the corpus downloaded, the following command will perform the complete regression, end to end, on any machine:

python src/main/python/run_regression.py --index --verify --search \
  --regression beir-v1.0.0-cqadupstack-webmasters-splade-distil-cocodenser-medium \
  --corpus-path collections/beir-v1.0.0-splade_distil_cocodenser_medium-cqadupstack-webmasters

Alternatively, you can simply copy/paste from the commands below and obtain the same results.

Indexing

Sample indexing command:

target/appassembler/bin/IndexCollection \
  -collection JsonVectorCollection \
  -input /path/to/beir-v1.0.0-cqadupstack-webmasters-splade_distil_cocodenser_medium \
  -generator DefaultLuceneDocumentGenerator \
  -index indexes/lucene-index.beir-v1.0.0-cqadupstack-webmasters-splade_distil_cocodenser_medium/ \
  -threads 16 -impact -pretokenized \
  >& logs/log.beir-v1.0.0-cqadupstack-webmasters-splade_distil_cocodenser_medium &

The path /path/to/beir-v1.0.0-cqadupstack-webmasters-splade_distil_cocodenser_medium/ should point to the corpus downloaded above.

The important indexing options to note here are -impact -pretokenized: the first tells Anserini not to encode BM25 doclengths into Lucene's norms (which is the default) and the second option says not to apply any additional tokenization on the pre-encoded tokens. Upon completion, we should have an index with 8,674 documents.

For additional details, see explanation of common indexing options.

Retrieval

Topics and qrels are stored here, which is linked to the Anserini repo as a submodule.

After indexing has completed, you should be able to perform retrieval as follows:

target/appassembler/bin/SearchCollection \
  -index indexes/lucene-index.beir-v1.0.0-cqadupstack-webmasters-splade_distil_cocodenser_medium/ \
  -topics tools/topics-and-qrels/topics.beir-v1.0.0-cqadupstack-webmasters.test.splade_distil_cocodenser_medium.tsv.gz \
  -topicReader TsvString \
  -output runs/run.beir-v1.0.0-cqadupstack-webmasters-splade_distil_cocodenser_medium.splade_distil_cocodenser_medium.topics.beir-v1.0.0-cqadupstack-webmasters.test.splade_distil_cocodenser_medium.txt \
  -impact -pretokenized -removeQuery -hits 1000 &

Evaluation can be performed using trec_eval:

tools/eval/trec_eval.9.0.4/trec_eval -c -m ndcg_cut.10 tools/topics-and-qrels/qrels.beir-v1.0.0-cqadupstack-webmasters.test.txt runs/run.beir-v1.0.0-cqadupstack-webmasters-splade_distil_cocodenser_medium.splade_distil_cocodenser_medium.topics.beir-v1.0.0-cqadupstack-webmasters.test.splade_distil_cocodenser_medium.txt
tools/eval/trec_eval.9.0.4/trec_eval -c -m recall.100 tools/topics-and-qrels/qrels.beir-v1.0.0-cqadupstack-webmasters.test.txt runs/run.beir-v1.0.0-cqadupstack-webmasters-splade_distil_cocodenser_medium.splade_distil_cocodenser_medium.topics.beir-v1.0.0-cqadupstack-webmasters.test.splade_distil_cocodenser_medium.txt
tools/eval/trec_eval.9.0.4/trec_eval -c -m recall.1000 tools/topics-and-qrels/qrels.beir-v1.0.0-cqadupstack-webmasters.test.txt runs/run.beir-v1.0.0-cqadupstack-webmasters-splade_distil_cocodenser_medium.splade_distil_cocodenser_medium.topics.beir-v1.0.0-cqadupstack-webmasters.test.splade_distil_cocodenser_medium.txt

Effectiveness

With the above commands, you should be able to reproduce the following results:

nDCG@10 SPLADE-distill CoCodenser Medium
BEIR (v1.0.0): CQADupStack-webmasters 0.3343
R@100 SPLADE-distill CoCodenser Medium
BEIR (v1.0.0): CQADupStack-webmasters 0.6404
R@1000 SPLADE-distill CoCodenser Medium
BEIR (v1.0.0): CQADupStack-webmasters 0.8767

Reproduction Log*

To add to this reproduction log, modify this template and run bin/build.sh to rebuild the documentation.