Skip to content

Add support for trec-cast v2 and v3 #294

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 2 commits into
base: master
Choose a base branch
from
Open

Add support for trec-cast v2 and v3 #294

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
0af4558
refactor trec_cast to not register subsets
andreaschari May 13, 2025
c62ac9a
Merge branch 'allenai:master' into master
andreaschari May 13, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
37 changes: 18 additions & 19 deletions ir_datasets/datasets/trec_cast.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -636,27 +636,26 @@ def wapo_converter(dsid, dupes: Dupes):
# marco_duplicates.txt for MS-MARCO

def register_docs(namespace: str, use_docs: bool, *tuples):
"""Register all documents (sub)collections
"""Registers documents (sub)collections

Tuples: (name prefix, document ID prefix, raw documents, passage count)
"""

all_docs_spec = []
all_passages_spec = []
passages = []

for dsid, prefix, raw, count in tuples:
prefixed = PrefixedDocs(None, PrefixedDocsSpec(prefix, raw))
subsets[f"{namespace}/{dsid}"] = Dataset(prefixed)

segmented = SegmentedDocs(
prefixed,
dlc[f"{namespace}/offsets/{dsid}"],
f"{NAME}/docs_{namespace}_{dsid}",
)
subsets[f"{namespace}/{dsid}/segmented"] = Dataset(segmented)

passage = CastPassageDocs(segmented, count)
passages.append(passage)
subsets[f"{namespace}/{dsid}/passages"] = Dataset(passage)

# Add this
all_docs_spec.append(
Expand All @@ -670,11 +669,11 @@ def register_docs(namespace: str, use_docs: bool, *tuples):
all_docs = PrefixedDocs(f"{NAME}/docs_{namespace}", *all_docs_spec)
subsets[f"{namespace}"] = Dataset(all_docs)
if use_docs:
# Add a passage dataset
# # Add a passage dataset
subsets[f"{namespace}/passages"] = PrefixedDocs(f"{NAME}/passages_{namespace}", *all_passages_spec)
return all_docs

"""

docs_v2 = register_docs(
"v2",
True,
Expand Down Expand Up @@ -704,20 +703,20 @@ def register_docs(namespace: str, use_docs: bool, *tuples):
),
("kilt", "KILT_", KiltCastDocs("kilt"), 17_124_025),
)
"""


#subsets["v2/2021"] = Dataset(
# docs_v2,
# CastQueries(dlc["2021/queries"], Cast2021Query),
# TrecQrels(dlc["2021/qrels"], QRELS_DEFS),
#)
subsets["v2/2021"] = Dataset(
docs_v2,
CastQueries(dlc["2021/queries"], Cast2021Query),
TrecQrels(dlc["2021/qrels"], QRELS_DEFS),
)

# --- Version 3 (2022)
# https://github.com/daltonj/treccastweb#year-4-trec-2022
# Official documents = processed (split) WAPO 2020, KILT, MS Marco V2

v3_dupes = dlc["v3/dupes"]
"""

docs_v3 = register_docs(
"v3",
False,
Expand Down Expand Up @@ -754,13 +753,13 @@ def register_docs(namespace: str, use_docs: bool, *tuples):
17_111_488,
),
)
"""


#subsets["v3/2022"] = Dataset(
# docs_v3,
# CastQueries(dlc["2022/queries"], Cast2022Query),
# TrecQrels(dlc["2022/qrels"], QRELS_DEFS),
#)
subsets["v3/2022"] = Dataset(
docs_v3,
CastQueries(dlc["2022/queries"], Cast2022Query),
TrecQrels(dlc["2022/qrels"], QRELS_DEFS),
)

# --- Register all datasets
ir_datasets.registry.register(NAME, base)
Expand Down