Create speech_flame_codes.py

speech_flame_codes.py

@@ -0,0 +1,84 @@
+# Lint as: python3
+"""semantic, acoustic and flame codes dataset.
+"""
+
+
+import glob
+import os
+
+import datasets
+import torch
+
+
+class SpeechFlameCodesDatasetConfig(datasets.BuilderConfig):
+    """BuilderConfig for Speech-Flame Codes dataset."""
+
+    def __init__(self, **kwargs):
+        super(SpeechFlameCodesDatasetConfig, self).__init__(**kwargs)
+
+
+class SpeechFlameCodesDataset(datasets.GeneratorBasedBuilder):
+    """Codes dataset."""
+
+    BUILDER_CONFIGS = [
+        SpeechFlameCodesDatasetConfig(name="all", description="SpeechFlameCodes dataset"),
+    ]
+
+    @property
+    def manual_download_instructions(self):
+        return (
+            "Codes should be computed before using this dataset. "
+            "`datasets.load_dataset('/path/to/this/script', name=all, data_dir='path/to/folder/folder_name/of/codes')`"
+        )
+
+    def _info(self):
+        features = datasets.Features(
+            {
+                "id": datasets.Value("string"),
+                "length": datasets.Value("int32"),
+                "acoustic_tokens": datasets.Array2D(shape=(None, 12), dtype="int16"),
+                "semantic_tokens": datasets.Array2D(shape=(None, 1), dtype="int16"),
+                "flame_tokens": datasets.Array2D(shape=(None, 1), dtype="int16"),
+            }
+        )
+
+        return datasets.DatasetInfo(
+            features=features,
+        )
+
+    def _split_generators(self, dl_manager):
+        base_data_dir = os.path.abspath(os.path.expanduser(dl_manager.manual_dir or ""))
+        if not os.path.exists(base_data_dir):
+            raise FileNotFoundError(
+                f"{base_data_dir} does not exist. Make sure you insert a manual dir via "
+                f"`datasets.load_dataset('/this/script', data_dir=...)` "
+                f"that includes code files .pt files "
+                f"dataset. Manual download instructions: {self.manual_download_instructions}"
+            )
+
+        train_data_dirs = glob.glob(os.path.join(base_data_dir, "*.pt"), recursive=False)
+        train_data_dirs = [d for d in train_data_dirs if '.ipynb_checkpoints' not in d]
+
+        return [
+            datasets.SplitGenerator(
+                name=str(datasets.Split.TRAIN),
+                gen_kwargs={"data_dirs": train_data_dirs},
+            ),
+        ]
+
+    def _generate_examples(self, data_dirs):
+        for key, path in enumerate(data_dirs):
+            id_ = path.split("/")[-1].replace(".pt", "")
+
+            data = torch.load(path, map_location="cpu")
+
+            acoustic_tokens = data["acoustic_codes"].transpose(0, 1)
+            semantic_tokens = data["semantic_codes"].unsqueeze(-1)
+            flame_tokens = data["flame_codes"].unsqueeze(-1)
+            
+            yield id_, {
+                "id": id_,
+                "acoustic_tokens": acoustic_tokens,
+                "semantic_tokens": semantic_tokens,
+                "flame_tokens": flame_tokens,
+            }