Skip to content

Latest commit

 

History

History
110 lines (78 loc) · 7.69 KB

regressions-msmarco-passage-splade-distil-cocodenser-medium.md

File metadata and controls

110 lines (78 loc) · 7.69 KB
Raw

Anserini Regressions: MS MARCO Passage Ranking

Model: SPLADE-distil CoCodenser Medium

This page describes regression experiments, integrated into Anserini's regression testing framework, using the SPLADE-distil CoCodenser Medium model on the MS MARCO passage ranking task. 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 msmarco-passage-splade-distil-cocodenser-medium

We make available a version of the MS MARCO Passage Corpus that has already been processed with SPLADE-distil CoCodenser Medium, i.e., performed model inference on every document and stored the output sparse vectors. Thus, no neural inference is involved.

From any machine, the following command will download the corpus and perform the complete regression, end to end:

python src/main/python/run_regression.py --download --index --verify --search --regression msmarco-passage-splade-distil-cocodenser-medium

The run_regression.py script automates the following steps, but if you want to perform each step manually, simply copy/paste from the commands below and you'll obtain the same regression results.

Corpus Download

Download the corpus and unpack into collections/:

wget https://rgw.cs.uwaterloo.ca/JIMMYLIN-bucket0/data/msmarco-passage-splade_distil_cocodenser_medium.tar -P collections/
tar xvf collections/msmarco-passage-splade_distil_cocodenser_medium.tar -C collections/

To confirm, msmarco-passage-splade_distil_cocodenser_medium.tar is 4.9 GB and has MD5 checksum f77239a26d08856e6491a34062893b0c. With the corpus downloaded, the following command will perform the remaining steps below:

python src/main/python/run_regression.py --index --verify --search --regression msmarco-passage-splade-distil-cocodenser-medium \
  --corpus-path collections/msmarco-passage-splade_distil_cocodenser_medium

Indexing

Sample indexing command:

target/appassembler/bin/IndexCollection \
  -collection JsonVectorCollection \
  -input /path/to/msmarco-passage-splade_distil_cocodenser_medium \
  -index indexes/lucene-index.msmarco-passage-splade_distil_cocodenser_medium/ \
  -generator DefaultLuceneDocumentGenerator \
  -threads 16 -impact -pretokenized -storeDocvectors \
  >& logs/log.msmarco-passage-splade_distil_cocodenser_medium &

The path /path/to/msmarco-passage-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 doc lengths 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 8,841,823 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. The regression experiments here evaluate on the 6980 dev set questions; see this page for more details.

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

target/appassembler/bin/SearchCollection \
  -index indexes/lucene-index.msmarco-passage-splade_distil_cocodenser_medium/ \
  -topics tools/topics-and-qrels/topics.msmarco-passage.dev-subset.splade_distil_cocodenser_medium.tsv.gz \
  -topicreader TsvInt \
  -output runs/run.msmarco-passage-splade_distil_cocodenser_medium.splade_distil_cocodenser_medium.topics.msmarco-passage.dev-subset.splade_distil_cocodenser_medium.txt \
  -impact -pretokenized &

Evaluation can be performed using trec_eval:

tools/eval/trec_eval.9.0.4/trec_eval -c -m map tools/topics-and-qrels/qrels.msmarco-passage.dev-subset.txt runs/run.msmarco-passage-splade_distil_cocodenser_medium.splade_distil_cocodenser_medium.topics.msmarco-passage.dev-subset.splade_distil_cocodenser_medium.txt
tools/eval/trec_eval.9.0.4/trec_eval -c -M 10 -m recip_rank tools/topics-and-qrels/qrels.msmarco-passage.dev-subset.txt runs/run.msmarco-passage-splade_distil_cocodenser_medium.splade_distil_cocodenser_medium.topics.msmarco-passage.dev-subset.splade_distil_cocodenser_medium.txt
tools/eval/trec_eval.9.0.4/trec_eval -c -m recall.100 tools/topics-and-qrels/qrels.msmarco-passage.dev-subset.txt runs/run.msmarco-passage-splade_distil_cocodenser_medium.splade_distil_cocodenser_medium.topics.msmarco-passage.dev-subset.splade_distil_cocodenser_medium.txt
tools/eval/trec_eval.9.0.4/trec_eval -c -m recall.1000 tools/topics-and-qrels/qrels.msmarco-passage.dev-subset.txt runs/run.msmarco-passage-splade_distil_cocodenser_medium.splade_distil_cocodenser_medium.topics.msmarco-passage.dev-subset.splade_distil_cocodenser_medium.txt

Effectiveness

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

AP@1000 SPLADE-distill CoCodenser Medium
MS MARCO Passage: Dev 0.3943
RR@10 SPLADE-distill CoCodenser Medium
MS MARCO Passage: Dev 0.3892
R@100 SPLADE-distill CoCodenser Medium
MS MARCO Passage: Dev 0.9111
R@1000 SPLADE-distill CoCodenser Medium
MS MARCO Passage: Dev 0.9817

Reproduction Log*

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