Jimmy Lin
Jimmy Lin
yea, this shouldn't happen... :(
which version of Pyserini are you depending on? probably an older version that doesn't have the `raw` method?
@sahel-sh see attached. [run.index.neuclir22-fa-en.test_title.bm25-default.gz](https://github.com/user-attachments/files/23343660/run.index.neuclir22-fa-en.test_title.bm25-default.gz)
Thanks @sahel-sh for the detailed analysis. Here's my proposal - - I've added a detailed comment in #2330 - Let's go ahead and merge #2330 - Let's leave this issue...
Interesting: the test cases _don't_ test the multi-metric case: https://github.com/castorini/pyserini/blob/master/tests/core/test_eval.py
good catch @lilyjge - that entire page needs an update. I'll work on that.
See https://github.com/castorini/pyserini/pull/1451 - does this help?
Unfortunately, the document deletion bindings have not been exposed on the Java end (from Lucene), so this is not currently doable. You're certainly welcome to send a PR to implement...
Hi @suraj-subrahmanyan thanks for working on this! Do the prebuilt indexes (e.g., https://github.com/castorini/anserini/blob/master/docs/fatjar-regressions/fatjar-regressions-v1.2.2.md ) still work? Or will we have to completely rebuild indexes?
hi @suraj-subrahmanyan I noticed that the HNSW scores change... which means we'll need to run regressions from scratch and manually verify/fix all scores... it's going to be a lot of...