refactored loading code + added sentence, pos tags and lemmas to data

data/wmt/dev/de.jsonl.bz2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dd141234e2b65337ccdf5305324bac1da5c8b4917b1689debb441b2567969b62
-size 169392
+oid sha256:eda8d6aeca141a8c34e032f659c8a892e6c77604300381a855264c72bb30473a
+size 367992

data/wmt/dev/en.jsonl.bz2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:55d27d06fa40208dfaf1919d8b874fbeae1b9a0ab06dd4cba167d2e22d3381ac
-size 183842
+oid sha256:147d0f564316c1d0780bd0af0a355d4656b74a19fa0352cee23dee42db05ff3c
+size 368183

data/wmt/test_2014/de.jsonl.bz2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ced62c34e465ce0eb67746531c91c5648b54d7372ccd750a14530c3a165dd1e0
-size 171597
+oid sha256:b94aa0ed447f0133dddc54498a42e73ee021bb81c89c8ecfd798bd65fd8066ce
+size 373427

data/wmt/test_2014/en.jsonl.bz2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:258d1906a0cb68f9939a8b74fe74041acc0254e5f831d4da26d3a1398c5ca553
-size 187684
+oid sha256:372c75cf8b18dbf0c64aa0c4ab4c93fb4bfa241052801f6ed5a42530b5963ff0
+size 374467

data/wmt/test_2019/de.jsonl.bz2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a682d8f42ac794118a5958f654b500c72e5a248d451b2be2438e29d558226717
-size 127100
+oid sha256:d70113241fabd3b6c8127ebbe3601e5f1e378615c176166c14e335277d5220d3
+size 273950

data/wmt/test_2019/en.jsonl.bz2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:466459f22abcb6acde1b1aed0c9c64cf36b60b42de2109488873a58453fd44ed
-size 131697
+oid sha256:b4f1cacf7555ac17cd9ce6ebac67b373215c3705db32c9c69bba78e7cab8624c
+size 264962

data/wmt/train/de.jsonl.bz2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4295b402a0a6c3d9c7b9034ba379555746f2b5dce1bc587d62083ca23d2f0e3f
-size 260858455
+oid sha256:ed11bc799e51fc0b0eafc03e2a01ddfdc606b36453679e0470167f37c8ee130e
+size 581073607

data/wmt/train/en.jsonl.bz2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7fb90965d5687574af9967e96511ae919726ef0b6fa407fafd5b0c1af61fb74e
-size 272337101
+oid sha256:f922f44e79e951910e407a8c0b449b1a53a05e32e34b8bb04ba10134184e1b09
+size 565677111

wsdmt.py

@@ -5,8 +5,13 @@ import datasets
 from datasets import DownloadManager, DatasetInfo
 
 
+def _order_langs(lang1, lang2):
+    return (lang1, lang2) if lang1 < lang2 else (lang2, lang1)
+
+
 class WSDMTConfig(datasets.BuilderConfig):
     def __init__(self, *args, corpus, lang1, lang2, **kwargs):
+        lang1, lang2 = _order_langs(lang1, lang2)
         super().__init__(
             *args,
             name=f"{corpus}@{lang1}-{lang2}",
@@ -20,6 +25,25 @@ class WSDMTConfig(datasets.BuilderConfig):
         return f"data/{self.corpus}/{split}/{lang}.jsonl.bz2"
 
 
+POS_TAGS = """ADJ
+ADP
+ADV
+AUX
+CCONJ
+DET
+INTJ
+NOUN
+NUM
+PART
+PRON
+PROPN
+PUNCT
+SCONJ
+SYM
+VERB
+X""".splitlines()
+
+
 class WSDMTDataset(datasets.GeneratorBasedBuilder):
     BUILDER_CONFIG_CLASS = WSDMTConfig
     config: WSDMTConfig
@@ -27,12 +51,8 @@ class WSDMTDataset(datasets.GeneratorBasedBuilder):
     def _generate_examples(self, path_lang1, path_lang2):
         with bz2.open(path_lang1) as f1, bz2.open(path_lang2) as f2:
             for n, (line1, line2) in enumerate(zip(f1, f2)):
-                sentence1_data = json.loads(line1)
-                sentence2_data = json.loads(line2)
-
-                texts1, senses1, is_senses1, is_polysemous1 = zip(*sentence1_data['data'])
-                texts2, senses2, is_senses2, is_polysemous2 = zip(*sentence2_data['data'])
-                sid1, sid2 = sentence1_data['sid'], sentence2_data['sid']
+                sid1, data1 = self._read_json_line(line1)
+                sid2, data2 = self._read_json_line(line2)
 
                 assert sid1 == sid2, (
                     f"Different sentence id found for {self.config.lang1} and {self.config.lang2}: "
@@ -41,34 +61,48 @@ class WSDMTDataset(datasets.GeneratorBasedBuilder):
 
                 data_dict = {
                     'sid': sid1,
-                    self.config.lang1: dict(tokens=texts1, sense=senses1,
-                                            identified_as_sense=is_senses1,
-                                            is_polysemous=is_polysemous1),
-                    self.config.lang2: dict(tokens=texts2, sense=senses2,
-                                            identified_as_sense=is_senses2,
-                                            is_polysemous=is_polysemous2),
+                    self.config.lang1: data1,
+                    self.config.lang2: data2,
                 }
 
                 yield n, data_dict
 
+    @classmethod
+    def _read_json_line(cls, line):
+        obj = json.loads(line)
+        sid = obj.pop('sid')
+        sentence = obj.pop('sentence')
+        data = obj.pop('data')
+        tokens, lemmas, pos_tags, senses, is_senses, is_polysemous = zip(*data)
+        assert len(tokens) == len(lemmas) == len(pos_tags) == len(senses) == len(is_senses) == len(is_polysemous), (
+            f"Inconsistent annotation lengths in sentence {sid}"
+        )
+
+        return sid, dict(
+            sentence=sentence,
+            tokens=tokens, lemmas=lemmas, pos_tags=pos_tags,
+            sense=senses, identified_as_sense=is_senses, is_polysemous=is_polysemous,
+        )
+
     def _info(self) -> DatasetInfo:
+        language_features = dict(
+            sentence=datasets.Value("string"),
+            tokens=datasets.Sequence(datasets.Value("string")),
+            sense=datasets.Sequence(datasets.Value("string")),
+            identified_as_sense=datasets.Sequence(datasets.Value("bool")),
+            is_polysemous=datasets.Sequence(datasets.Value("bool")),
+            lemmas=datasets.Sequence(datasets.Value("string")),
+            pos_tags=datasets.Sequence(datasets.ClassLabel(names=POS_TAGS)),
+            # pos_tags=datasets.Sequence(datasets.Value("string")),
+        )
+
         return datasets.DatasetInfo(
             description="empty description",
             features=datasets.Features(
                 {
                     "sid": datasets.Value("string"),
-                    self.config.lang1: {
-                        "tokens": datasets.Sequence(datasets.Value("string")),
-                        "sense": datasets.Sequence(datasets.Value("string")),
-                        "identified_as_sense": datasets.Sequence(datasets.Value("bool")),
-                        "is_polysemous": datasets.Sequence(datasets.Value("bool")),
-                    },
-                    self.config.lang2: {
-                        "tokens": datasets.Sequence(datasets.Value("string")),
-                        "sense": datasets.Sequence(datasets.Value("string")),
-                        "identified_as_sense": datasets.Sequence(datasets.Value("bool")),
-                        "is_polysemous": datasets.Sequence(datasets.Value("bool")),
-                    }
+                    self.config.lang1: language_features,
+                    self.config.lang2: language_features
                 },
             ),
             supervised_keys=None,