Massive v1.1 loader

README.md

@@ -11,6 +11,7 @@ multilinguality:
 - ar-SA
 - az-AZ
 - bn-BD
+- ca-ES
 - cy-GB
 - da-DK
 - de-DE
@@ -190,13 +191,14 @@ The dataset can be used to train a model for `natural-language-understanding` (N
 
 ### Languages
 
-The corpora consists of parallel sentences from 51 languages :
+The massive 1.1 corpora consists of parallel sentences from 52 languages :
 
 - `Afrikaans - South Africa (af-ZA)`
 - `Amharic - Ethiopia (am-ET)`
 - `Arabic - Saudi Arabia (ar-SA)`
 - `Azeri - Azerbaijan (az-AZ)`
 - `Bengali - Bangladesh (bn-BD)`
+- `Catalan - Spain (ca-ES)`
 - `Chinese - China (zh-CN)`
 - `Chinese - Taiwan (zh-TW)`
 - `Danish - Denmark (da-DK)`

massive.py

@@ -441,16 +441,22 @@ licenses.
 
 Creative Commons may be contacted at creativecommons.org.
 """
+# version 1.0
+_URL0 = "https://amazon-massive-nlu-dataset.s3.amazonaws.com/amazon-massive-dataset-1.0.tar.gz"
+# version 1.1
+_URL1 = "https://amazon-massive-nlu-dataset.s3.amazonaws.com/amazon-massive-dataset-1.1.tar.gz"
 
-_URL = "https://amazon-massive-nlu-dataset.s3.amazonaws.com/amazon-massive-dataset-1.0.tar.gz"
-
-_LANGUAGES = ['af-ZA', 'am-ET', 'ar-SA', 'az-AZ', 'bn-BD', 'cy-GB', 'da-DK', 'de-DE', 'el-GR', 'en-US',
+# version 1.0
+_LANGUAGES0 = ['af-ZA', 'am-ET', 'ar-SA', 'az-AZ', 'bn-BD', 'cy-GB', 'da-DK', 'de-DE', 'el-GR', 'en-US',
              'es-ES', 'fa-IR', 'fi-FI', 'fr-FR', 'he-IL', 'hi-IN', 'hu-HU', 'hy-AM', 'id-ID', 'is-IS',
              'it-IT', 'ja-JP', 'jv-ID', 'ka-GE', 'km-KH', 'kn-IN', 'ko-KR', 'lv-LV', 'ml-IN', 'mn-MN',
              'ms-MY', 'my-MM', 'nb-NO', 'nl-NL', 'pl-PL', 'pt-PT', 'ro-RO', 'ru-RU', 'sl-SL', 'sq-AL',
              'sv-SE', 'sw-KE', 'ta-IN', 'te-IN', 'th-TH', 'tl-PH', 'tr-TR', 'ur-PK', 'vi-VN', 'zh-CN',
              'zh-TW']
 
+# version 1.1 ('ca-ES' added)
+_LANGUAGES1 = sorted(_LANGUAGES0 + ['ca-ES'])
+
 _SCENARIOS = ['social', 'transport', 'calendar', 'play', 'news', 'datetime', 'recommendation', 'email',
               'iot', 'general', 'audio', 'lists', 'qa', 'cooking', 'takeaway', 'music', 'alarm', 'weather']
 
@@ -470,22 +476,48 @@ _INTENTS = ['datetime_query', 'iot_hue_lightchange', 'transport_ticket', 'takeaw
 
 _ALL = "all"
 
+
+class MassiveConfig(datasets.BuilderConfig):
+    """BuilderConfig for MASSIVE."""
+
+    def __init__(self, dataset_version=None, *args, **kwargs):
+        """BuilderConfig for MASSIVE.
+        Args:
+          **kwargs: keyword arguments forwarded to super.
+        """
+        super(MassiveConfig, self).__init__(*args, **kwargs)
+        self.dataset_version = dataset_version if dataset_version else "1.0"
+        self.data_url = _URL0 if self.dataset_version == "1.0" else _URL1
+
+
 class MASSIVE(datasets.GeneratorBasedBuilder):
     """MASSIVE: A 1M-Example Multilingual Natural Language Understanding Dataset with 51 Typologically-Diverse Languages"""
 
+    # All individual locale datasets are served from the latest version.
     BUILDER_CONFIGS = [
-        datasets.BuilderConfig(
+        MassiveConfig(
             name = name,
+            dataset_version = '1.1',
             version = datasets.Version("1.0.0"),
-            description = f"The MASSIVE corpora for {name}",
-        ) for name in _LANGUAGES
-    ]
-
-    BUILDER_CONFIGS.append(datasets.BuilderConfig(
-        name = _ALL,
-        version = datasets.Version("1.0.0"),
-        description = f"The MASSIVE corpora for entire corpus",
-    ))
+            description = f"The MASSIVE v1.1 corpora for {name}",
+        ) for name in _LANGUAGES1
+        ]
+    # Version 1.0
+    BUILDER_CONFIGS.append(MassiveConfig(
+            name = _ALL,
+            dataset_version = '1.0',
+            version = datasets.Version("1.0.0"),
+            description = f"The MASSIVE v1.0 corpora for entire corpus",
+        )
+    )
+    # Version 1.1
+    BUILDER_CONFIGS.append(MassiveConfig(
+            name = _ALL + '_1.1', # Append '_1.1' to 'all' in order to get the latest version.
+            dataset_version = '1.1',
+            version = datasets.Version("1.0.0"),
+            description = f"The MASSIVE v1.1 corpora for entire corpus",
+        )
+    )
 
     DEFAULT_CONFIG_NAME = _ALL
 
@@ -524,8 +556,7 @@ class MASSIVE(datasets.GeneratorBasedBuilder):
 
     def _split_generators(self, dl_manager):
 
-        archive = dl_manager.download(_URL)
-
+        archive = dl_manager.download(self.config.data_url)
         return [
             datasets.SplitGenerator(
                 name=datasets.Split.TRAIN,
@@ -557,8 +588,10 @@ class MASSIVE(datasets.GeneratorBasedBuilder):
 
         key_ = 0
 
-        if lang == "all":
-            lang = _LANGUAGES.copy()
+        if lang == "all_1.1":
+            lang = _LANGUAGES1.copy()
+        elif lang == "all":
+            lang = _LANGUAGES0.copy()
         else:
             lang = [lang]
 
@@ -566,7 +599,7 @@ class MASSIVE(datasets.GeneratorBasedBuilder):
 
         for path, f in files:
 
-            l = path.split("1.0/data/")[-1].split(".jsonl")[0]
+            l = path.split("/")[-1].split(".")[0]
 
             if not lang:
                 break

test_massive.py

@@ -2,6 +2,17 @@ from datasets import load_dataset
 
 source = "AmazonScience/massive"
 
+#Individual locales are downloaded from version 1.1 which has Catalan (ca-ES) added in
 dataset = load_dataset(source, "en-US", download_mode="force_redownload")
 print(dataset)
 print(dataset["train"][0])
+
+#Using version 1.1 which has Catalan (ca-ES) added in
+dataset = load_dataset(source, "all_v1.1", download_mode="force_redownload")
+print(dataset)
+print(dataset["train"][0])
+
+#Using version 1.0
+dataset = load_dataset(source, "all", download_mode="force_redownload")
+print(dataset)
+print(dataset["train"][0])