trying again by getting rid of dataset class script

README.md

@@ -5,28 +5,14 @@ task_categories:
 language:
 - pt
 pretty_name: LapsBenchmark
-annotations_creators:
-- found
-language_creators:
-- found
 multilinguality:
 - monolingual
 size_categories:
 - 10K<n<100K
-source_datasets:
-- original
-dataset_info:
-  features:
-  - name: audio
-    dtype:
-      audio:
-        sampling_rate: 16000
-  - name: text
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 3621471
-    num_examples: 20
-  download_size: 3739136
-  dataset_size: 3621471
+configs:
+- config_name: default
+  data_files:
+  - split: test
+    path:
+    - "data/LapsBM-F004.tar"
 ---

dataset.py

@@ -1,62 +0,0 @@
-import io
-import os
-import tarfile
-
-import datasets
-import soundfile as sf
-
-
-class LapsBenchmarkDataset(datasets.GeneratorBasedBuilder):
-    BUILDER_CONFIGS = [
-        datasets.BuilderConfig(name="default", version=datasets.Version("1.0.0"))
-    ]
-    DEFAULT_CONFIG_NAME = "default"
-
-    def _info(self):
-        return datasets.DatasetInfo(
-            features=datasets.Features({
-                "audio": datasets.Audio(sampling_rate=16_000),
-                "text": datasets.Value("string")
-            }),
-        )
-
-    def _split_generators(self, dl_manager):
-        archive_path = dl_manager.download("data/LapsBM-F004.tar")
-        return [
-                datasets.SplitGenerator(
-                    name=datasets.Split.TRAIN,
-                    gen_kwargs={"archive_path": archive_path}
-                )
-            ]
-
-    def _generate_examples(self, archive_path):
-        with tarfile.open(archive_path) as tar:
-            wav, txt = {}, {}
-            for member in tar.getmembers():
-                uttid, _ = os.path.splitext(os.path.basename(member.name))
-                if not os.path.basename(member.name).startswith("LapsBM"):
-                    continue
-                if member.name.endswith(".wav"):
-                    audio_bytes = tar.extractfile(member).read()
-                    audio_array, _ = sf.read(io.BytesIO(audio_bytes))
-                    wav[uttid] = audio_array
-                elif member.name.endswith(".txt"):
-                    text_bytes = tar.extractfile(member).read()
-                    text_string = text_bytes.decode("utf-8").strip()
-                    txt[uttid] = text_string
-            for uttid in sorted(set(list(wav.keys()) + list(txt.keys()))):
-                samples = wav[uttid]
-                text = txt[uttid]
-                yield uttid, {
-                    "audio": {
-                        "array": samples,
-                        "sampling_rate": 16000,
-                    },
-                    "text": text
-                }
-
-
-if __name__ == "__main__":
-    d = LapsBenchmarkDataset()
-    egs = d._generate_examples("data/LapsBM-F004.tar")
-    print(next(egs))