fix: Compatibility with huggingface_hub 1.16, numpy 2.0, and pandas 3.0 (#1971)

* fix: Use namespaced HF dataset paths for huggingface_hub >= 1.16 compatibility

huggingface_hub 1.16+ enforces that dataset repository IDs must use the
'namespace/name' format. Bare dataset names (e.g., 'hellaswag') are no
longer accepted, causing HfUriError in CI.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: Yoav Katz <katz@il.ibm.com>

* fix: Regenerate catalog JSONs with namespaced HF dataset paths

Run all prepare/cards scripts to update the catalog JSON files with the
full namespace/name format for HuggingFace dataset paths.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: Yoav Katz <katz@il.ibm.com>

* fix: Replace evaluate library with direct sklearn calls in F1MultiLabel

The evaluate library's cached f1.py uses `float(score)` on numpy arrays,
which raises TypeError with numpy >= 2.0. Bypass the evaluate wrapper and
call sklearn's f1_score/precision_score/recall_score directly.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: Yoav Katz <katz@il.ibm.com>

* fix: Numpy 2.0 and pandas 3.0 compatibility fixes

- F1 metric: replace evaluate library wrapper with direct sklearn calls
  to avoid numpy 2.0 float() TypeError on 0-d arrays
- text2sql: cast DataFrame to str before sorting to avoid pandas 3.0
  TypeError when assigning string values to int64 columns
- wiki_bio: use namespaced HF dataset path (michaelauli/wiki_bio)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: Yoav Katz <katz@il.ibm.com>

* fix: Regenerate wiki_bio catalog JSON with namespaced HF path

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: Yoav Katz <katz@il.ibm.com>

* fix: Use namespaced HF path for winogrande (allenai/winogrande)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: Yoav Katz <katz@il.ibm.com>

---------

Signed-off-by: Yoav Katz <katz@il.ibm.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

Author: yoavkatz

prepare/cards/arc.py

@@ -7,7 +7,7 @@
 
 for subtask in subtasks:
     card = TaskCard(
-        loader=LoadHF(path="ai2_arc", name=subtask),
+        loader=LoadHF(path="allenai/ai2_arc", name=subtask),
         preprocess_steps=[
             Set({"topic": "science"}),
             Rename(field_to_field={"answerKey": "label", "choices": "_choices"}),

prepare/cards/billsum.py

@@ -10,7 +10,7 @@
 n_chars_to_filter_by_list = ["max", 6000, 10000]
 for n_chars_to_filter_by in n_chars_to_filter_by_list:
     card = TaskCard(
-        loader=LoadHF(path="billsum"),
+        loader=LoadHF(path="FiscalNote/billsum"),
         preprocess_steps=[
             SplitRandomMix(
                 {"train": "train[87.5%]", "validation": "train[12.5%]", "test": "test"}

prepare/cards/clinc_oos.py

@@ -171,7 +171,7 @@
 
 for subset in ["small", "imbalanced", "plus"]:
     card = TaskCard(
-        loader=LoadHF(path="clinc_oos", name=subset),
+        loader=LoadHF(path="clinc/clinc_oos", name=subset),
         preprocess_steps=[
             Shuffle(page_size=sys.maxsize),
             Rename(field_to_field={"intent": "label"}),

prepare/cards/cnn_dailymail.py

@@ -9,7 +9,7 @@
 from unitxt.test_utils.card import test_card
 
 card = TaskCard(
-    loader=LoadHF(path="cnn_dailymail", name="3.0.0"),
+    loader=LoadHF(path="abisee/cnn_dailymail", name="3.0.0"),
     preprocess_steps=[
         Rename(field_to_field={"article": "document"}),
         Wrap(field="highlights", inside="list", to_field="summaries"),

prepare/cards/copa.py

@@ -10,7 +10,7 @@
 from unitxt.test_utils.card import test_card
 
 card = TaskCard(
-    loader=LoadHF(path="super_glue", name="copa"),
+    loader=LoadHF(path="aps/super_glue", name="copa"),
     preprocess_steps=[
         "splitters.small_no_test",
         ListFieldValues(fields=["choice1", "choice2"], to_field="choices"),

prepare/cards/dbpedia_14.py

@@ -34,7 +34,7 @@
 mappers = {str(i): cls for i, cls in enumerate(classes)}
 
 card = TaskCard(
-    loader=LoadHF(path=f"{dataset_name}"),
+    loader=LoadHF(path="fancyzhx/dbpedia_14"),
     preprocess_steps=[
         Shuffle(page_size=sys.maxsize),
         SplitRandomMix(

prepare/cards/ethos.py

@@ -12,7 +12,7 @@
 
 card = TaskCard(
     loader=LoadHF(
-        path="ethos",
+        path="iamollas/ethos",
         revision="refs/convert/parquet",
         data_dir="binary",
         splits=["train"],

prepare/cards/go_emotions.py

@@ -11,13 +11,13 @@
 dataset_name = "go_emotions"
 subset = "simplified"
 
-ds_builder = load_dataset_builder(dataset_name, subset)
+ds_builder = load_dataset_builder("google-research-datasets/go_emotions", subset)
 classes = ds_builder.info.features["labels"].feature.names
 
 mappers = {str(i): cls for i, cls in enumerate(classes)}
 
 card = TaskCard(
-    loader=LoadHF(path=dataset_name, name=subset),
+    loader=LoadHF(path="google-research-datasets/go_emotions", name=subset),
     preprocess_steps=[
         MapInstanceValues(mappers={"labels": mappers}, process_every_value=True),
         Set(

prepare/cards/hellaswag.py

@@ -4,7 +4,7 @@
 from unitxt.test_utils.card import test_card
 
 card = TaskCard(
-    loader=LoadHF(path="hellaswag"),
+    loader=LoadHF(path="Rowan/hellaswag"),
     preprocess_steps=[
         "splitters.large_no_test",
         Rename(

prepare/cards/human_eval.py

@@ -15,7 +15,7 @@
     get_asserts = '[t for t in re.findall(r"assert.*?(?=\\n\\s*assert|$)", test.replace("candidate", entry_point), re.DOTALL)]'
 
     card = TaskCard(
-        loader=LoadHF(path="openai_humaneval", split="test"),
+        loader=LoadHF(path="openai/openai_humaneval", split="test"),
         preprocess_steps=[
             ExecuteExpression(
                 expression=get_asserts, imports_list=["re"], to_field="test_list"