Fix bird dataset (#1593)

add streaming, add tests

examples/evaluate_text2sql.py

@@ -37,7 +37,7 @@
     evaluated_dataset[0]["score"]["global"]["score"] >= 0.44
 ), "results have been degraded, something is wrong with the metric"
 
-# with llama-3-70b-instruct
+# with llama-3-3-70b-instruct
 # num_of_instances (int):
 #     100
 # execution_accuracy (float):

prepare/cards/text2sql.py

@@ -4,9 +4,10 @@
 from unitxt.blocks import Copy, Rename, Set, TaskCard
 from unitxt.loaders import LoadHF
 from unitxt.operators import ExecuteExpression, Shuffle
+from unitxt.test_utils.card import test_card
 
 card = TaskCard(
-    loader=LoadHF(path="premai-io/birdbench", split="validation"),
+    loader=LoadHF(path="premai-io/birdbench", split="validation", streaming=True),
     preprocess_steps=[
         Shuffle(page_size=sys.maxsize),
         Rename(
@@ -45,9 +46,15 @@
 )
 
 
-# test_card(
-#     card, num_demos=0, demos_pool_size=0, demos_taken_from="test"
-# )
+test_card(
+    card,
+    num_demos=0,
+    demos_pool_size=0,
+    demos_taken_from="validation",
+    loader_limit=10,
+    test_exact_match_score_when_predictions_equal_references=False,
+    test_full_mismatch_score_with_full_mismatch_prediction_values=False,
+)
 
 add_to_catalog(
     card,

src/unitxt/catalog/cards/text2sql/bird.json

@@ -3,7 +3,8 @@
     "loader": {
         "__type__": "load_hf",
         "path": "premai-io/birdbench",
-        "split": "validation"
+        "split": "validation",
+        "streaming": true
     },
     "preprocess_steps": [
         {