datamaestro-text 2025.4.3__py3-none-any.whl → 2025.6.11__py3-none-any.whl

datamaestro_text/config/com/github/aagohary/canard.py

@@ -1,7 +1,5 @@
-# See documentation on https://datamaestro.readthedocs.io
-
 from datamaestro.definitions import datatasks, datatags, dataset
-from datamaestro.download.archive import zipdownloader
+from datamaestro.download.single import filedownloader
 from datamaestro.utils import HashCheck
 
 from datamaestro.data.ml import Supervised
@@ -10,14 +8,23 @@ from datamaestro_text.data.conversation.canard import CanardDataset
 
 @datatags("conversation", "context", "query")
 @datatasks("query rewriting")
-@zipdownloader(
-    "archive",
-    "https://obj.umiacs.umd.edu/elgohary/CANARD_Release.zip",
-    subpath="CANARD_Release",
-    checker=HashCheck("c9bba7c6bb898f669383415b54fd6ffd"),
+@filedownloader(
+    "train.json",
+    "https://raw.githubusercontent.com/aagohary/canard/refs/heads/master/data/release/train.json",
+    checker=HashCheck("73624ac646fb81e09b0fd7f01370ada3"),
+)
+@filedownloader(
+    "dev.json",
+    "https://raw.githubusercontent.com/aagohary/canard/refs/heads/master/data/release/dev.json",
+    checker=HashCheck("c84525631a83bc771c58ff31f4a9b601"),
+)
+@filedownloader(
+    "test.json",
+    "https://raw.githubusercontent.com/aagohary/canard/refs/heads/master/data/release/test.json",
+    checker=HashCheck("3fc14d0078e7a5056f5da571728f024e"),
 )
 @dataset(Supervised, url="https://sites.google.com/view/qanta/projects/canard", id="")
-def main(archive):
+def main(train, dev, test):
     """Question-in-context rewriting
 
     CANARD is a dataset for question-in-context rewriting that consists of
@@ -30,7 +37,7 @@ def main(archive):
     Each dataset is an instance of :class:`datamaestro_text.data.conversation.CanardDataset`
     """
     return {
-        "train": CanardDataset(path=archive / "train.json"),
-        "validation": CanardDataset(path=archive / "dev.json"),
-        "test": CanardDataset(path=archive / "test.json"),
+        "train": CanardDataset(path=train),
+        "validation": CanardDataset(path=dev),
+        "test": CanardDataset(path=test),
     }

datamaestro_text/config/com/github/apple/ml-qrecc.py

@@ -25,12 +25,11 @@ from datamaestro_text.datasets.irds.helpers import lz4docstore_builder
     checker=HashCheck("f88fcc7ef3678cd6312080389c8abd67"),
 )
 @dataset(
-    Supervised[QReCCDataset, None, QReCCDataset],
     url="https://github.com/apple/ml-qrecc",
     doi="https://doi.org/10.48550/arXiv.2010.04898",
     id="",
 )
-def main(data: Path):
+def main(data: Path) -> Supervised[QReCCDataset, None, QReCCDataset]:
     """Open-Domain Question Answering Goes Conversational via Question Rewriting
 
     We introduce QReCC (Question Rewriting in Conversational Context), an
@@ -40,10 +39,10 @@ def main(data: Path):
     answering that includes the individual subtasks of question rewriting,
     passage retrieval and reading comprehension
     """
-    return {
-        "train": QReCCDataset(path=data / "qrecc_train.json"),
-        "test": QReCCDataset(path=data / "qrecc_test.json"),
-    }
+    return Supervised(
+        train=QReCCDataset(path=data / "qrecc_train.json"),
+        test=QReCCDataset(path=data / "qrecc_test.json"),
+    )
 
 
 @dataset(
@@ -52,7 +51,6 @@ def main(data: Path):
 )
 class Content(LZ4JSONLDocumentStore):
     """QReCC mentionned URLs content"""
-
     @staticmethod
     def __create_dataset__(dataset, options=None):
         ds = reference(reference=main).setup(dataset, options)
@@ -67,7 +65,7 @@ class Content(LZ4JSONLDocumentStore):
             "id",
         ).setup(dataset, options)
 
-        return LZ4JSONLDocumentStore(jsonl_path=store_path)
+        return Content(jsonl_path=store_path)
 
     @staticmethod
     def _documents(path: Path):

datamaestro_text/config/com/github/prdwb/orconvqa.py

@@ -49,9 +49,9 @@ def preprocessed(train, dev, test):
     Each dataset is an instance of :class:`datamaestro_text.data.conversation.OrConvQADataset`
     """
     return {
-        "train": OrConvQADataset(path=train),
-        "validation": OrConvQADataset(path=dev),
-        "test": OrConvQADataset(path=test),
+        "train": OrConvQADataset.C(path=train),
+        "validation": OrConvQADataset.C(path=dev),
+        "test": OrConvQADataset.C(path=test),
     }
 
 

datamaestro_text/config/com/sentiment140.py

@@ -1,11 +1,10 @@
 from datamaestro.data.csv import Generic
-from datamaestro.definitions import argument, datatasks, datatags, dataset
+from datamaestro.definitions import datatasks, datatags, dataset
 from datamaestro.download.archive import zipdownloader
 from datamaestro.data.ml import Supervised
 from datamaestro.utils import HashCheck
 
 
-
 @zipdownloader(
     "dir",
     "http://cs.stanford.edu/people/alecmgo/trainingandtestdata.zip",

datamaestro_text/config/edu/stanford/glove.py

@@ -11,6 +11,7 @@ from datamaestro.download.archive import zipdownloader
 from datamaestro.download.single import filedownloader
 from datamaestro_text.data.embeddings import WordEmbeddingsText
 
+
 # size: 822M
 # statistics:
 #   tokens: 6G

datamaestro_text/data/conversation/base.py

@@ -129,6 +129,8 @@ class ConversationNode:
 
 
 class ConversationTree(ABC):
+    """Represents a conversation tree"""
+
     @abstractmethod
     def root(self) -> ConversationNode:
         ...
@@ -192,7 +194,7 @@ class SingleConversationTreeNode(ConversationNode):
         return (
             SingleConversationTreeNode(self.tree, self.index + 1)
             if self.index < len(self.tree.history) - 1
-            else []
+            else None
         )
 
     def children(self) -> List[ConversationNode]:
@@ -207,8 +209,8 @@ class ConversationTreeNode(ConversationNode, ConversationTree):
     """A conversation tree node"""
 
     entry: Record
-    parent: Optional["ConversationTreeNode"]
-    children: List["ConversationTreeNode"]
+    _parent: Optional["ConversationTreeNode"]
+    _children: List["ConversationTreeNode"]
 
     def __init__(self, entry):
         self.entry = entry
@@ -216,8 +218,8 @@ class ConversationTreeNode(ConversationNode, ConversationTree):
         self.children = []
 
     def add(self, node: "ConversationTreeNode") -> "ConversationTreeNode":
-        self.children.append(node)
-        node.parent = self
+        self._children.append(node)
+        node._parent = self
         return node
 
     def conversation(self, skip_self: bool) -> ConversationHistory:
@@ -232,14 +234,14 @@ class ConversationTreeNode(ConversationNode, ConversationTree):
     def __iter__(self) -> Iterator["ConversationTreeNode"]:
         """Iterates over all conversation tree nodes (pre-order)"""
         yield self.entry
-        for child in self.children:
+        for child in self._children:
             yield from child
 
     def parent(self) -> Optional[ConversationNode]:
-        return self.parent
+        return self._parent
 
     def children(self) -> List[ConversationNode]:
-        return self.children
+        return self._children
 
     def root(self):
         return self

datamaestro_text/data/conversation/canard.py

@@ -1,12 +1,17 @@
 from typing import Iterator, List
 from attr import define
 import json
+from datamaestro.record import Record
 from datamaestro.data import File
-from .base import (
+from datamaestro_text.data.conversation.base import (
+    ConversationDataset,
     ConversationTree,
     SingleConversationTree,
+    SimpleDecontextualizedItem,
+    EntryType,
 )
-from . import ConversationDataset
+from datamaestro_text.data.ir import IDItem, SimpleTextItem
+import logging
 
 
 @define(kw_only=True)
@@ -30,7 +35,10 @@ class CanardConversation:
 
 
 class CanardDataset(ConversationDataset, File):
-    """A dataset in the CANARD JSON format"""
+    """A dataset in the CANARD JSON format
+
+    The CANARD dataset is composed of
+    """
 
     def entries(self) -> Iterator[CanardConversation]:
         """Iterates over re-written query with their context"""
@@ -47,22 +55,53 @@ class CanardDataset(ConversationDataset, File):
             )
 
     def __iter__(self) -> Iterator[ConversationTree]:
-        history = []
+        history: list[Record] = []
         current_id = None
 
         for entry in self.entries():
-            # Check if current conversation
-            if current_id != entry.dialogue_id and current_id is not None:
-                history.reverse()
-                yield SingleConversationTree(current_id, history)
+            # Check if current conversation, otherwise we are OK
+            if current_id != entry.dialogue_id:
+                if current_id is not None:
+                    history.reverse()
+                    yield SingleConversationTree(current_id, history)
+                    history = []
+
+                current_id = entry.dialogue_id
+
+            if not history:
+                # First round
+                # The two first items are the wikipedia title and section,
+                # we interpret them as two user queries
+                assert len(entry.history) == 2
+                history.extend(
+                    Record(
+                        SimpleTextItem(text),
+                        EntryType.USER_QUERY,
+                    )
+                    for text in entry.history
+                )
+            else:
+                # The utterance before the last is the last user query
+                assert (
+                    entry.history[-2] == history[-1][SimpleTextItem].text
+                ), f"{entry.dialogue_id} {entry.history} / {history[-4:-1]}"
+
+                # The last utterance is the system side
+                history.append(
+                    Record(SimpleTextItem(entry.history[-1]), EntryType.SYSTEM_ANSWER)
+                )
+
+            assert len(entry.history) == len(history)
 
             # Add to current
             history.append(
-                # FIXME: not working anymore
-                CanardEntry(
-                    query=entry.query,
-                    decontextualized_query=entry.rewrite,
+                Record(
+                    IDItem(f"{entry.dialogue_id}-{entry.query_no}"),
+                    SimpleTextItem(entry.query),
+                    SimpleDecontextualizedItem(entry.rewrite),
+                    EntryType.USER_QUERY,
                 )
             )
 
-        yield current
+        if current_id:
+            yield SingleConversationTree(current_id, history)

datamaestro_text/data/conversation/orconvqa.py

@@ -1,4 +1,3 @@
-from functools import cached_property
 from typing import Iterator, List, Optional
 from attr import define
 import json

datamaestro_text/data/ir/base.py

@@ -43,6 +43,12 @@ class IDItem(Item, ABC):
 
     id: str
 
+@define
+class UrlItem(Item):
+    """An url item"""
+
+    url: str
+
 
 @define
 class AdhocAssessment:

datamaestro_text/data/ir/formats.py

@@ -58,7 +58,7 @@ class NFCorpusDocument(TextItem):
 
     @cached_property
     def text(self):
-        return self.abstract
+        return f"{self.title} {self.abstract}"
 
 
 @define
@@ -99,8 +99,8 @@ class WapoDocument(TextItem):
     body_media: Tuple[WapoDocMedia, ...]
 
     @cached_property
-    def text(self):
-        return self.body
+    def text(self):        
+        return f"{self.title} {self.body_paras_html}"
 
 
 @define
@@ -126,13 +126,24 @@ class OrConvQADocument(TextItem):
     def text(self):
         return f"{self.title} {self.body}"
 
+
 @define
+class DprW100Doc(TextItem):
+    text: str
+    title: str
+
+@define 
+class MsMarcoV2Passage(TextItem):
+    text: str
+    spans: Tuple[Tuple[int, int], ...]
+    msmarco_document_id: str
 class Touche2020(TextItem):
     text: str
     title: str
     stance: str
     url: str
 
+
 @define
 class SciDocs(TextItem):
     text: str
@@ -175,7 +186,8 @@ class TrecMb14Query(TextItem):
     def get_text(self):
         return f"{self.query}"
 
-@define 
+
+@define
 class SciDocsTopic(TextItem):
     text: str
     authors: List[str]
@@ -183,6 +195,7 @@ class SciDocsTopic(TextItem):
     cited_by: List[str]
     references: List[str]
 
+
 @define()
 class TrecTopic(SimpleTextItem):
     description: str
@@ -190,3 +203,17 @@ class TrecTopic(SimpleTextItem):
 
 
 TrecTopicRecord = record_type(IDItem, TrecTopic)
+
+@define
+class DprW100Query(TextItem):
+    text: str
+    answers: Tuple[str]
+
+@define
+class TrecBackgroundLinkingQuery(IDItem):
+    query_id: str
+    doc_id: str
+    url: str
+
+    def get_text(self):
+        raise NotImplementedError()

datamaestro_text/datasets/irds/data.py

@@ -37,6 +37,7 @@ from datamaestro_text.data.ir.base import (
     SimpleAdhocAssessment,
     SimpleTextItem,
     TopicRecord,
+    UrlItem,
     create_record,
 )
 
@@ -165,6 +166,19 @@ class Documents(ir.DocumentStore, IRDSId):
             "source",
             "source_content_type",
         ),
+        _irds.dpr_w100.DprW100Doc: tuple_constructor(
+            formats.DprW100Doc,
+            "doc_id",
+            "text",
+            "title",
+        ),
+        _irds.msmarco_passage_v2.MsMarcoV2Passage: tuple_constructor(
+            formats.MsMarcoV2Passage,
+            "doc_id",
+            "text",
+            "spans",
+            "msmarco_document_id",
+        ),
     }
 
     """Wraps an ir datasets collection -- and provide a default text
@@ -385,6 +399,12 @@ class Topics(ir.TopicsStore, IRDSId):
             "tweet_time",
             "description",
         ),
+        _irds.dpr_w100.DprW100Query: tuple_constructor(
+            formats.DprW100Query,
+            "query_id",
+            "text",
+            "answers"
+        ),
     }
 
     HANDLERS = {
@@ -415,7 +435,52 @@ class Topics(ir.TopicsStore, IRDSId):
     def iter(self) -> Iterator[TopicRecord]:
         """Returns an iterator over topics"""
         return self.handler.iter()
+    
+class TrecBackgroundLinkingTopicsHandler(TopicsHandler):
+    def __init__(self, dataset):
+        self.dataset = dataset
+    
+    @cached_property
+    def ext2records(self):
+        return {record[IDItem].id: record for record in self.records}
+
+    def topic_int(self, internal_topic_id: int) -> TopicRecord:
+        """Returns a document given its internal ID"""
+        return self.records[internal_topic_id]
+
+    def topic_ext(self, external_topic_id: str) -> TopicRecord:
+        """Returns a document given its external ID"""
+        return self.ext2records[external_topic_id]
 
+    def iter(self) -> Iterator[ir.TopicRecord]:
+        """Returns an iterator over topics"""
+        return iter(self.records)
+
+    @cached_property
+    def records(self):
+        try:
+            records = []
+
+            for query in self.dataset.dataset.queries_iter():
+                topic =  Record(
+                    IDItem(query.query_id),
+                    # Following BEIR documentation, we use title of documents as queries: https://github.com/beir-cellar/beir/blob/main/examples/dataset/README.md#queries-and-qrels
+                    SimpleTextItem(self.dataset.dataset.docs_store().get(query.doc_id).title), 
+                    UrlItem(query.url),
+                )
+                records.append(topic)
+        except Exception:
+            logging.exception("Error while computing topic records")
+            raise
+
+        return records
+
+
+Topics.HANDLERS.update(
+    {
+        _irds.wapo.TrecBackgroundLinkingQuery: TrecBackgroundLinkingTopicsHandler
+    }
+)
 
 class CastTopicsHandler(TopicsHandler):
     def __init__(self, dataset):

datamaestro_text/datasets/irds/datasets.py

@@ -65,7 +65,7 @@ class QrelsDataset(Dataset):
         return True
 
     def _prepare(self, download=False) -> Documents:
-        return AdhocAssessments(id=self.fullid)
+        return AdhocAssessments.C(id=self.fullid)
 
 
 class QueriesDataset(Dataset):
@@ -78,7 +78,7 @@ class QueriesDataset(Dataset):
         return True
 
     def _prepare(self, download=False) -> Documents:
-        return Topics(id=self.fullid)
+        return Topics.C(id=self.fullid)
 
 
 # class ScoredDocuments(Dataset):
@@ -96,7 +96,7 @@ class DocumentsDataset(Dataset):
         return True
 
     def _prepare(self, download=False) -> Documents:
-        return Documents(id=self.fullid)
+        return Documents.C(id=self.fullid)
 
 
 class TrainingTripletsDataset(Dataset):
@@ -116,10 +116,6 @@ class AdhocRunDataset(Dataset):
     def _prepare(self, download=False) -> AdhocRun:
         return AdhocRun(id=self.fullid)
 
-    @property
-    def configtype(self):
-        return AdhocRun
-
 
 class Collection(Dataset):
     base = Adhoc
@@ -127,7 +123,7 @@ class Collection(Dataset):
     topics: QueriesDataset
 
     def _prepare(self, download=False) -> Documents:
-        return Adhoc(
+        return Adhoc.C(
             id=self.fullid,
             topics=self.topics.prepare(download),
             assessments=self.assessments.prepare(download),

datamaestro_text/utils/iter.py

@@ -1,4 +1,4 @@
-from typing import Callable, TypeVar, Iterator, List, Union
+from typing import Callable, Sequence, TypeVar, Iterator, List, Union
 
 T = TypeVar("T")
 
@@ -45,7 +45,7 @@ class RangeView:
         return RangeView(self.source, key)
 
 
-class LazyList:
+class LazyList(Sequence):
     """Iterable-based list
 
     The list is only materialized if needed"""
@@ -63,6 +63,9 @@ class LazyList:
         else:
             return iter(self.materialized_list)
 
+    def __len__(self):
+        return len(self.iterable)
+
     def __getitem__(self, index):
         # Materialize the list if accessing an index above the threshold or any slice
         if isinstance(index, slice) or index >= self.materialize_threshold:

datamaestro_text/version.py

@@ -17,5 +17,5 @@ __version__: str
 __version_tuple__: VERSION_TUPLE
 version_tuple: VERSION_TUPLE
 
-__version__ = version = '2025.4.3'
-__version_tuple__ = version_tuple = (2025, 4, 3)
+__version__ = version = '2025.6.11'
+__version_tuple__ = version_tuple = (2025, 6, 11)

{datamaestro_text-2025.4.3.dist-info → datamaestro_text-2025.6.11.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: datamaestro-text
-Version: 2025.4.3
+Version: 2025.6.11
 Summary: Datamaestro module for text-related datasets
 Author-email: Benjamin Piwowarski <benjamin@piwowarski.fr>
 License: GPL-3
@@ -18,9 +18,15 @@ Classifier: Topic :: Software Development :: Libraries :: Python Modules
 Requires-Python: >=3.8
 Description-Content-Type: text/markdown
 License-File: LICENSE
-Requires-Dist: datamaestro>=1.2.1
+Requires-Dist: datamaestro>=1.4.2
 Requires-Dist: ir_datasets>=0.5.8
 Requires-Dist: attrs
+Provides-Extra: dev
+Requires-Dist: pytest; extra == "dev"
+Requires-Dist: docutils; extra == "dev"
+Requires-Dist: sphobjinv; extra == "dev"
+Requires-Dist: flake8; extra == "dev"
+Requires-Dist: sphinx; extra == "dev"
 Dynamic: license-file
 
  [![pre-commit](https://img.shields.io/badge/pre--commit-enabled-brightgreen?logo=pre-commit&logoColor=white)](https://github.com/pre-commit/pre-commit) [![PyPI version](https://badge.fury.io/py/datamaestro-text.svg)](https://badge.fury.io/py/datamaestro-text)

{datamaestro_text-2025.4.3.dist-info → datamaestro_text-2025.6.11.dist-info}/RECORD

@@ -1,13 +1,13 @@
 datamaestro_text/__init__.py,sha256=hU8jZpkXl3F74qIfqnJl7v4nJ9YxfoR7IpJpUREFNRI,248
-datamaestro_text/version.py,sha256=Ldmqy5wkUM54W7PO84xMGLTubji7Xl68QRaeaSSILS0,517
+datamaestro_text/version.py,sha256=M2FaCl2nFdTTxwIhOsBo1SQ_3ytid7NHYp2QLIimPXY,519
 datamaestro_text/config/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 datamaestro_text/config/ai/quac.yaml,sha256=h1D7UJo2z1nZ_9MXpDHuQNJG9Ma2oayUdJV6lyumAIg,1103
 datamaestro_text/config/com/oscar-corpus.py,sha256=dDfdy_uhG6esRQQzUOOORwK8KvEq4c4hZbWt22pv-zY,733
-datamaestro_text/config/com/sentiment140.py,sha256=bLxFY6xIOp3_9mn5H36V-jfa_vXdetRxi6sK4cghl9w,1294
+datamaestro_text/config/com/sentiment140.py,sha256=3cZfqs395gY14gGojRC_RTYEdG5l8RUTSdF9li8ynBc,1283
 datamaestro_text/config/com/fastml/goodbooks-10k.yaml,sha256=5ZABxUnBFs2ZnCXtBH8YoBiPb3SocRRdH1DLSfVWF-Y,1172
-datamaestro_text/config/com/github/aagohary/canard.py,sha256=tIwb_KxMUR3st7rzQUkt6rIjolTl8uKvDq6t795b1nY,1468
-datamaestro_text/config/com/github/apple/ml-qrecc.py,sha256=_lufJlg-4zaQyjr5Ae-X-9hXzpl2JFjfxh_RhnBbva4,3068
-datamaestro_text/config/com/github/prdwb/orconvqa.py,sha256=Tg3hbnvilxW_Lwt7fpKvGMtu-6mc9oNIHM-LX6JTR1U,3026
+datamaestro_text/config/com/github/aagohary/canard.py,sha256=DREIKU2FFrfwibjU48Y1AsKTUzhwxmD4Kuykd4bDIus,1762
+datamaestro_text/config/com/github/apple/ml-qrecc.py,sha256=-iHKJocWZ8N9N-P8E45y4ewg3OT_23XonlDh5-NcH2g,3055
+datamaestro_text/config/com/github/prdwb/orconvqa.py,sha256=MFJYaxJoqJQ6kMDHa2PIE0zoxYTA8Kyl26-vzFoMML0,3032
 datamaestro_text/config/com/github/soskek/bookcorpus.yaml,sha256=qJKs35yeEIilEMgNvU3OEqMp1TSn7mDM2T-uYyA7kTU,1607
 datamaestro_text/config/com/microsoft/wikiqa.yaml,sha256=U7rU-W2Xz1MYv2YXT4jCTj5DsHyM0VssbQPNa3EISaM,540
 datamaestro_text/config/com/microsoft/msmarco/passage.py,sha256=tVldwVh0pQpXXifqcMNEN9cLO1HXkXoEhToSazSx5RE,11643
@@ -15,7 +15,7 @@ datamaestro_text/config/com/smashwords/bookcorpus.py,sha256=hCFjZg9t1AHET05qU31u
 datamaestro_text/config/edu/cornell/nlvr.yaml,sha256=9Yk5VZMncSmrP7JNuGXqExksgX5nQ_Zfnlps8hWze3Q,921
 datamaestro_text/config/edu/stanford/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 datamaestro_text/config/edu/stanford/aclimdb.py,sha256=lsPDxnp_rWOCpBte6pZ0_LVaC33w5mmgfGh51rcTgt8,643
-datamaestro_text/config/edu/stanford/glove.py,sha256=ykkQ7nYWqhmgc2TeohNMliYSiX831cYUygftkBTGIac,2390
+datamaestro_text/config/edu/stanford/glove.py,sha256=bXxwiJqT8alPs5MwwxVuY0xBO8g1QWhm2L3AKVskTlA,2391
 datamaestro_text/config/edu/stanford/im2p.yaml,sha256=JoToNyEPpmwdyLFedCBot5ypyw7p9rzi12mGXJuZin0,2909
 datamaestro_text/config/edu/upenn/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 datamaestro_text/config/edu/upenn/ldc/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -46,23 +46,23 @@ datamaestro_text/data/recommendation.py,sha256=wHV_9SXSclouuXaBmvwg3ncZLOFfdKRZG
 datamaestro_text/data/tagging.py,sha256=yWm7bNLks77cAySa1aZNCmLFxTTqhqXZ0PaoaEYU6hI,697
 datamaestro_text/data/text.py,sha256=_9J7-j06BOIZ1HsOcBsWy_asanAKkDoZIsWWMMj4tm4,496
 datamaestro_text/data/conversation/__init__.py,sha256=esOWnSImMlQs5QtfxUt559ABLd6a5wwoNpj9XtIq71Y,159
-datamaestro_text/data/conversation/base.py,sha256=zXX5sPMoVtDf4WTkLx45IvZz6lgKPgcUZFu6N4lVBlc,6457
-datamaestro_text/data/conversation/canard.py,sha256=IMxu5NfytWZLZ_cyT8UFOICbDE82HLJoNBT_6j36Faw,1808
-datamaestro_text/data/conversation/orconvqa.py,sha256=VG3GV5_IPQcvla9rrQPypDNcZYmKNMgozmr2oudeLA4,3802
+datamaestro_text/data/conversation/base.py,sha256=PUVRCSMBlV9bSayBl-vnzsYvyr2Tdv_zTadIC_Tswe0,6508
+datamaestro_text/data/conversation/canard.py,sha256=FI02rPNuzSsA_uXOcgcuQn31d3QahmHndzYCV1Zm8rk,3305
+datamaestro_text/data/conversation/orconvqa.py,sha256=RL9cpK7QK2dsk_j2e4Th8wzGFq-G3ww_EjdM1eKh-8o,3764
 datamaestro_text/data/conversation/qrecc.py,sha256=es4GmqPtE63A7O_GARe8Zy3rQvuLEhAvUA7CfN_nMeA,2562
 datamaestro_text/data/ir/__init__.py,sha256=EOVnRv9oKEEucMOgrFjhXxeUWEkaJkbXBndpkKhZaiY,8701
-datamaestro_text/data/ir/base.py,sha256=Cw8egjChpx4ksUwp-vTA70B2OWxROH7FIeJylsXggAk,1429
+datamaestro_text/data/ir/base.py,sha256=TSaY8UWoixqvKg1z5JfkXPnuZiVoChYEUvc7jjbpZqY,1495
 datamaestro_text/data/ir/cord19.py,sha256=yu1Khgy6AZjH2BPQKdnxDid0vQgQ8zvb8-FQlHH-GOU,1465
 datamaestro_text/data/ir/csv.py,sha256=tnxuqR_MZ3GQhuoXEMYyWLQw8PyD0gRqsnzIP5Gsziw,1212
 datamaestro_text/data/ir/data.py,sha256=ERmmOxz_9mUkIuccNbzUa5Y6gVLLVDdyc4cCxbCCUbY,20
-datamaestro_text/data/ir/formats.py,sha256=rU9uJhdFDdlnQ2qfFowK_--1WVOH1fVgfXWOPPDF_FY,3160
+datamaestro_text/data/ir/formats.py,sha256=K0mqmCGg0seneKo-Rt3cBfDVjEMS4_6t1MfNj4iW9Y4,3637
 datamaestro_text/data/ir/huggingface.py,sha256=G71VFDN-SllZy4LFxumEbCumEJvb5-orAbLemHjWhiA,956
 datamaestro_text/data/ir/stores.py,sha256=rlOogoBAfsP7o01KqvHiTF1XqzK2Fp6QbRmuasabKAE,859
 datamaestro_text/data/ir/trec.py,sha256=qDOzmgUn0hMqeP9KdI190-9IKsLl6yNRqIoBz6s-cBs,1898
 datamaestro_text/data/ir/utils.py,sha256=6-GhXVtgkBZGhIs2-ODZua_3DmKjSSVydStpHDqbAwE,833
 datamaestro_text/datasets/irds/__init__.py,sha256=Tq0HN1qojnZYLBumM59BuTkz7r0gcu-5OXmDDLgPpAc,707
-datamaestro_text/datasets/irds/data.py,sha256=BTsFZdjFJ0XdlYRh2rcshAoeknHrypf7be2BGUbWWFk,19931
-datamaestro_text/datasets/irds/datasets.py,sha256=qtN-nX2_OV9FD339aZjwVL1jFf6I4T6MbNCcuRAjmtU,5682
+datamaestro_text/datasets/irds/data.py,sha256=0V5nPMQeh1I3Sp36x_NE7xrbC_vbZAKhU5NONj95aok,22058
+datamaestro_text/datasets/irds/datasets.py,sha256=n5mvD2omVI4PkeuJgX13Cio6gLrgYjxq8ZUKKyqocEs,5625
 datamaestro_text/datasets/irds/helpers.py,sha256=fGE-fbuJbXdTzl1qo55247jzn9cvApY-d82GJBgfY1E,3982
 datamaestro_text/datasets/irds/utils.py,sha256=m30JLIrV_HgilN11TvY9dGTyumES6LLzWZDUAMT915M,1425
 datamaestro_text/download/tmdb.py,sha256=kU_Vz9jhznlyeKMHziVu58IHoWv8zPu6CZTHVNQvmu4,4009
@@ -75,12 +75,12 @@ datamaestro_text/transforms/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMp
 datamaestro_text/transforms/ir/__init__.py,sha256=eWxr0ZtfK9F9J4TachW5V8Ra5AhQJueDyOGR8vXi1uA,6553
 datamaestro_text/utils/__init__.py,sha256=2449YLTAtKJzkmt84Mu8sBRCCveNs5fiaqTCK_p5ha0,3340
 datamaestro_text/utils/files.py,sha256=n6ZGl5LNrZbHLcV9RFwd7cFT0vPUezit-2dsBzs1vRQ,170
-datamaestro_text/utils/iter.py,sha256=2_UZ8y9Ma4k5U9ZD4w55Zfb6NGrKM1L4G40OygRm1is,2459
+datamaestro_text/utils/iter.py,sha256=uzBmavBeqVyEvYmi9Ds2dqh0ywhJE-1SBO5v8AoIKcc,2537
 datamaestro_text/utils/randomstream.py,sha256=_-boH4IIqN8qcl3IktjpNp9vmF4TWRzHUSNVwg7WAr8,973
 datamaestro_text/utils/shuffle.py,sha256=o8JTz3mr0lYWyv0zEh91jEK12ci1etMiUnzh5GkOHCM,3490
-datamaestro_text-2025.4.3.dist-info/licenses/LICENSE,sha256=OXLcl0T2SZ8Pmy2_dmlvKuetivmyPd5m1q-Gyd-zaYY,35149
-datamaestro_text-2025.4.3.dist-info/METADATA,sha256=M1XG19GB1RLCTJ0xICe47LYDjHzLGFPUvHXg9-bmZZM,1631
-datamaestro_text-2025.4.3.dist-info/WHEEL,sha256=CmyFI0kx5cdEMTLiONQRbGQwjIoR1aIYB7eCAQ4KPJ0,91
-datamaestro_text-2025.4.3.dist-info/entry_points.txt,sha256=lO1P5hE183L5qEEVHlG8d_ik0HNXnX7Eo87cQLdcl-Y,111
-datamaestro_text-2025.4.3.dist-info/top_level.txt,sha256=gYSeqViE8r7eCxSdqFJL74OwljOwKsGPaIhEcCXqc-o,17
-datamaestro_text-2025.4.3.dist-info/RECORD,,
+datamaestro_text-2025.6.11.dist-info/licenses/LICENSE,sha256=OXLcl0T2SZ8Pmy2_dmlvKuetivmyPd5m1q-Gyd-zaYY,35149
+datamaestro_text-2025.6.11.dist-info/METADATA,sha256=oMXEU9_JPRCrUCoSE6IGmmyLzfdpVQgKHAaaEbFXMXY,1847
+datamaestro_text-2025.6.11.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+datamaestro_text-2025.6.11.dist-info/entry_points.txt,sha256=lO1P5hE183L5qEEVHlG8d_ik0HNXnX7Eo87cQLdcl-Y,111
+datamaestro_text-2025.6.11.dist-info/top_level.txt,sha256=gYSeqViE8r7eCxSdqFJL74OwljOwKsGPaIhEcCXqc-o,17
+datamaestro_text-2025.6.11.dist-info/RECORD,,

{datamaestro_text-2025.4.3.dist-info → datamaestro_text-2025.6.11.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (78.1.0)
+Generator: setuptools (80.9.0)
 Root-Is-Purelib: true
 Tag: py3-none-any