Jimmy Lin

Switch to Maven for build

1

Switch from Ant/Ivy for build and package management.

I've trained a model, and now I want to run inference over held out data. This doesn't seem to work...

Add ability to compute trec_eval metrics directly on in-memory data structures

4

From you paper: This is now pretty easy... with [pyserini on PyPI](https://pypi.org/project/pyserini/). But the real point of this issue is this: currently, as a I understand it, the input to...

In `~/.cache/pyserini/indexes/`, for pre-built indexes, we have: ``` lucene-index.msmarco-v1-passage.20221004.252b5e lucene-index.msmarco-v1-passage.20221004.252b5e.c697b18c9a0686ca760583e615dbe450 ``` When Pyserini downloads it, the directory name gets postpended with the MD5 checksum; this doesn't happen with Anserini, so...

We're still on JDK11. Should we jump all the way to JDK 17 or JDK 21? These are the LTS choices: https://www.oracle.com/ca-en/java/technologies/java-se-support-roadmap.html

Currently, for `trec_eval`, we have to do something like: ``` target/appassembler/bin/trec_eval -c -M 10 -m recip_rank tools/topics-and-qrels/qrels.msmarco-passage.dev-subset.txt runs/run.msmarco-passage.bm25.txt ``` It would be great if we could do something like: ```...

Errors with openai-ada2-int8 regressions: GCLocker errors

1

Here: https://github.com/castorini/anserini/blob/master/docs/regressions/regressions-msmarco-passage-openai-ada2-int8.md ``` nohup python src/main/python/run_regression.py --index --verify --search \ --regression msmarco-passage-openai-ada2-int8 \ --corpus-path collections/msmarco-passage-openai-ada2 >& log.openai-ada2-int8 & ``` Errors: ``` 2023-12-20 16:12:12,873 INFO [python] ========== Indexing ========== 2023-12-20 16:12:12,874...

Basic rank fusion implementation in Anserini

1

In order to fully repro #2307 in Anserini, we'll need some basic rank fusion impl in Anserini.

SearchCollection -rf.qrels option

1

Is the `-rf.qrels` in `SearchCollection` actually being used by any regression, external program, etc.?

Missing msmarco-doc-segmented-wp.yaml condition

1

```bash $ ls src/main/resources/regression/ | grep wp | grep -v beir dl19-doc-hgf-wp.yaml dl19-doc-segmented-wp.yaml dl19-doc-wp.yaml dl19-passage-hgf-wp.yaml dl19-passage-wp.yaml dl20-doc-hgf-wp.yaml dl20-doc-segmented-wp.yaml dl20-doc-wp.yaml dl20-passage-hgf-wp.yaml dl20-passage-wp.yaml msmarco-doc-hgf-wp.yaml msmarco-doc-segmented-wp.yaml msmarco-doc-wp.yaml msmarco-passage-hgf-wp.yaml msmarco-passage-wp.yaml ``` We appear to...