store local paths in non-streaming mode

peoples_speech.py

@@ -13,6 +13,7 @@
 # limitations under the License.
 
 import json
+import os
 
 import datasets
 from datasets.tasks import AutomaticSpeechRecognition
@@ -86,13 +87,13 @@ _URLS = {
     },
 }
 
-# _BASE_URL = "https://huggingface.co/datasets/MLCommons/peoples_speech/resolve/main/"
+_BASE_URL = "https://huggingface.co/datasets/MLCommons/peoples_speech/resolve/main/"
 
 # relative path to data inside dataset's repo
-_DATA_URL = "{config}/{config}_00000{archive_id}.tar"
+_DATA_URL = _BASE_URL + "{config}/{config}_00000{archive_id}.tar"
 
 # relative path to metadata inside dataset's repo
-_MANIFEST_URL = "{config}.json"
+_MANIFEST_URL = _BASE_URL + "{config}.json"
 
 
 class PeoplesSpeech(datasets.GeneratorBasedBuilder):
@@ -130,7 +131,9 @@ class PeoplesSpeech(datasets.GeneratorBasedBuilder):
         # TODO: for demo purposes I use just first 5 archives
         # TODO: this should be changed to the actual number of archives further
         urls = [_DATA_URL.format(config=self.config.name, archive_id=i) for i in range(5)]
-        archives = [dl_manager.iter_archive(dl_manager.download(url)) for url in urls]
+        archive_paths = [dl_manager.download(url) for url in urls]
+        local_extracted_archive_paths = [dl_manager.extract(path) for path in archive_paths] \
+            if not dl_manager.is_streaming else [None] * len(archive_paths)
 
         manifest_url = _MANIFEST_URL.format(config=self.config.name)
         manifest_path = dl_manager.download_and_extract(manifest_url)
@@ -139,13 +142,14 @@ class PeoplesSpeech(datasets.GeneratorBasedBuilder):
             datasets.SplitGenerator(
                 name=datasets.Split.TRAIN,
                 gen_kwargs={
-                    "archives": archives,
+                    "local_extracted_archive_paths": local_extracted_archive_paths,
+                    "archives": [dl_manager.iter_archive(path) for path in archive_paths],
                     "manifest_path": manifest_path
                 },
             ),
         ]
 
-    def _generate_examples(self, archives, manifest_path):
+    def _generate_examples(self, local_extracted_archive_paths, archives, manifest_path):
         meta = dict()
         with open(manifest_path, "r", encoding="utf-8") as f:
             for line in tqdm(f, desc="reading metadata file"):
@@ -162,13 +166,13 @@ class PeoplesSpeech(datasets.GeneratorBasedBuilder):
                     }
 
         print("generating examples")
-        for archive in archives:
-            # note that you don't need to use `tarfile` library and open tar archives manually
-            # dl_manager.iter_archive() does it for you :)
+        for local_extracted_archive_path, archive in zip(local_extracted_archive_paths, archives):
             for audio_filename, audio_file in archive:
+                path = os.path.join(local_extracted_archive_path, audio_filename) if local_extracted_archive_path \
+                    else audio_filename
                 yield audio_filename, {
                     "id": audio_filename,
-                    "audio": {"path": audio_filename, "bytes": audio_file.read()},
+                    "audio": {"path": path, "bytes": audio_file.read()},
                     "text": meta[audio_filename]["text"],
                     "duration_ms": meta[audio_filename]["duration_ms"]
                 }