up

ami-ihm.py

@@ -265,9 +265,9 @@ _AUDIO_ARCHIVE_URL = _BASE_DATA_URL + "audio/{subset}/{split}/{_id}.tar.gz"
 
 _ANNOTATIONS_ARCHIVE_URL = _BASE_DATA_URL + "annotations/{split}/text"
 
-_TRANSCRIPT_URL = "https://huggingface.co/datasets/distil-whisper/ami/resolve/main/transcription_data/greedy_search/"
+_TRANSCRIPT_URL = "https://huggingface.co/datasets/distil-whisper/ami-ihm/resolve/main/transcription_data/greedy_search/"
 
-_TRANSCRIPT_URLS = _TRANSCRIPT_URL + "{config}/{split}.txt"
+_TRANSCRIPT_URLS = _TRANSCRIPT_URL + "/{split}-transcription.txt"
 
 logger = datasets.utils.logging.get_logger(__name__)
 
@@ -314,6 +314,7 @@ class AMI(datasets.GeneratorBasedBuilder):
 
     def _split_generators(self, dl_manager):
         splits = ["train", "dev", "eval"]
+        splits_alt = ["train", "validation", "test"]
 
         audio_archives_urls = {}
         for split in splits:
@@ -334,8 +335,8 @@ class AMI(datasets.GeneratorBasedBuilder):
         annotations_urls = {split: _ANNOTATIONS_ARCHIVE_URL.format(split=split) for split in splits}
         annotations = dl_manager.download(annotations_urls)
 
-        transcription_urls = {split: _TRANSCRIPT_URLS.format(split=split) for split in splits}
-        transcript_archive_path = dl_manager.download(transcription_urls[self.config.name])
+        transcription_urls = {split: _TRANSCRIPT_URLS.format(split=split) for split in splits_alt}
+        transcript_archive_path = dl_manager.download(transcription_urls)
 
         return [
             datasets.SplitGenerator(
@@ -345,7 +346,8 @@ class AMI(datasets.GeneratorBasedBuilder):
                     "local_extracted_archives_paths": local_extracted_archives_paths["train"],
                     "annotation": annotations["train"],
                     "transcript_files": transcript_archive_path["train"],
-                    "split": "train"
+                    "split": "train",
+                    "config": self.config.name,
                 },
             ),
             datasets.SplitGenerator(
@@ -354,8 +356,9 @@ class AMI(datasets.GeneratorBasedBuilder):
                     "audio_archives": [dl_manager.iter_archive(archive) for archive in audio_archives["dev"]],
                     "local_extracted_archives_paths": local_extracted_archives_paths["dev"],
                     "annotation": annotations["dev"],
-                    "transcript_files": transcript_archive_path["dev"],
-                    "split": "dev"
+                    "transcript_files": transcript_archive_path["validation"],
+                    "split": "dev",
+                    "config": self.config.name,
                 },
             ),
             datasets.SplitGenerator(
@@ -364,16 +367,20 @@ class AMI(datasets.GeneratorBasedBuilder):
                     "audio_archives": [dl_manager.iter_archive(archive) for archive in audio_archives["eval"]],
                     "local_extracted_archives_paths": local_extracted_archives_paths["eval"],
                     "annotation": annotations["eval"],
-                    "transcript_files": transcript_archive_path["eval"],
-                    "split": "eval"
+                    "transcript_files": transcript_archive_path["test"],
+                    "split": "eval",
+                    "config": self.config.name,
                 },
             ),
         ]
 
-    def _generate_examples(self, audio_archives, local_extracted_archives_paths, annotation, transcript_files, split):
+    def _generate_examples(self, audio_archives, local_extracted_archives_paths, annotation, transcript_files, split, config):
         # open annotation file
         assert len(audio_archives) == len(local_extracted_archives_paths)
 
+        if config != "ihm":
+            raise ValueError("This dataset is intended for the ihm split only. Use `distil-whisper/ami-sdm` for the sdm split")
+
         with open(annotation, "r", encoding="utf-8") as f:
             transcriptions = {}
             for line in f.readlines():