init

tweet_temporal_shift.py

@@ -2,7 +2,7 @@
 import json
 import datasets
 
-_VERSION = "0.0.1"
+_VERSION = "0.0.2"
 _TWEET_TEMPORAL_DESCRIPTION = """"""
 _TWEET_TEMPORAL_CITATION = """"""
 _TWEET_TOPIC_DESCRIPTION = """
@@ -117,13 +117,13 @@ class TweetTemporalShift(datasets.GeneratorBasedBuilder):
             features=["text", "gold_label_list", "date"],
             data_url=f"{_ROOT_URL}/tweet_topic",
         ),
-        TweetTemporalShiftConfig(
-            name="ner_temporal",
-            description=_TWEET_NER7_DESCRIPTION,
-            citation=_TWEET_NER7_CITATION,
-            features=["text", "text_tokenized", "gold_label_sequence", "date"],
-            data_url=f"{_ROOT_URL}/tweet_ner",
-        ),
+        # TweetTemporalShiftConfig(
+        #     name="ner_temporal",
+        #     description=_TWEET_NER7_DESCRIPTION,
+        #     citation=_TWEET_NER7_CITATION,
+        #     features=["text", "text_tokenized", "gold_label_sequence", "date"],
+        #     data_url=f"{_ROOT_URL}/tweet_ner",
+        # ),
         TweetTemporalShiftConfig(
             name="nerd_temporal",
             description=_TWEET_NERD_DESCRIPTION,
@@ -142,13 +142,13 @@ class TweetTemporalShift(datasets.GeneratorBasedBuilder):
                     features=["text", "gold_label_list", "date"],
                     data_url=f"{_ROOT_URL}/tweet_topic_test{i}_seed{s}",
                 ),
-                TweetTemporalShiftConfig(
-                    name=f"ner_random{i}_seed{s}",
-                    description=_TWEET_NER7_DESCRIPTION,
-                    citation=_TWEET_NER7_CITATION,
-                    features=["text", "text_tokenized", "gold_label_sequence", "date"],
-                    data_url=f"{_ROOT_URL}/tweet_ner_test{i}_seed{s}",
-                ),
+                # TweetTemporalShiftConfig(
+                #     name=f"ner_random{i}_seed{s}",
+                #     description=_TWEET_NER7_DESCRIPTION,
+                #     citation=_TWEET_NER7_CITATION,
+                #     features=["text", "text_tokenized", "gold_label_sequence", "date"],
+                #     data_url=f"{_ROOT_URL}/tweet_ner_test{i}_seed{s}",
+                # ),
                 TweetTemporalShiftConfig(
                     name=f"nerd_random{i}_seed{s}",
                     description=_TWEET_NERD_DESCRIPTION,
@@ -160,7 +160,7 @@ class TweetTemporalShift(datasets.GeneratorBasedBuilder):
 
     def _info(self):
         features = {feature: datasets.Value("string") for feature in self.config.features}
-        if "topic_" in self.config.name:
+        if "topic" in self.config.name:
             names = [
                 "arts_&_culture", "business_&_entrepreneurs", "celebrity_&_pop_culture", "diaries_&_daily_life",
                 "family", "fashion_&_style", "film_tv_&_video", "fitness_&_health", "food_&_dining", "gaming",
@@ -168,16 +168,7 @@ class TweetTemporalShift(datasets.GeneratorBasedBuilder):
                 "science_&_technology", "sports", "travel_&_adventure", "youth_&_student_life"]
             features["gold_label_list"] = datasets.Sequence(
                 datasets.features.ClassLabel(names=names))
-        if "ner_" in self.config.name:
-            names = [
-                "B-corporation", "B-creative_work", "B-event", "B-group", "B-location", "B-person", "B-product",
-                "I-corporation", "I-creative_work", "I-event", "I-group", "I-location", "I-person", "I-product", "O"]
-            features["gold_label_sequence"] = datasets.Sequence(
-                datasets.features.ClassLabel(names=names))
-            features["text_tokenized"] = datasets.Sequence(
-                datasets.Value("string"))
-            features["entities"] = datasets.features.Sequence({"entity": datasets.Value("string"), "type": datasets.Value("string")})
-        if "nerd_" in self.config.name:
+        elif "nerd" in self.config.name:
             features["target"] = datasets.Value("string")
             features["text"] = datasets.Value("string")
             features["definition"] = datasets.Value("string")
@@ -185,6 +176,15 @@ class TweetTemporalShift(datasets.GeneratorBasedBuilder):
             features["text_end"] = datasets.Value("int32")
             features["gold_label_binary"] = datasets.Value("int32")
             features["date"] = datasets.Value("string")
+        # elif "ner" in self.config.name:
+        #     names = [
+        #         "B-corporation", "B-creative_work", "B-event", "B-group", "B-location", "B-person", "B-product",
+        #         "I-corporation", "I-creative_work", "I-event", "I-group", "I-location", "I-person", "I-product", "O"]
+        #     features["gold_label_sequence"] = datasets.Sequence(datasets.features.ClassLabel(names=names))
+        #     features["text_tokenized"] = datasets.Sequence(datasets.Value("string"))
+        #     features["entities"] = datasets.features.Sequence(
+        #         {"entity": datasets.Value("string"), "type": datasets.Value("string")}
+        #     )
         return datasets.DatasetInfo(
             description=_TWEET_TEMPORAL_DESCRIPTION + "\n" + self.config.description,
             features=datasets.Features(features),