file ids - transcriptions

peoples_speech-clean.py

@@ -11,7 +11,7 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-
+import csv
 import json
 import os
 
@@ -72,9 +72,9 @@ _N_SHARDS_URL = _BASE_URL + "n_shards.json"
 # relative path to metadata inside dataset's repo
 _MANIFEST_URL = _BASE_URL + "{split}/{config}.json"
 
-_WHISPER_TRANSCRIPT_URL = "https://huggingface.co/datasets/distil-whisper/peoples_speech-clean/resolve/main/transcription_data/greedy_search/"
+_WHISPER_TRANSCRIPT_URL = "https://huggingface.co/datasets/distil-whisper/whisper_transcriptions_greedy/resolve/main/peoples_speech-clean"
 
-_WHISPER_TRANSCRIPT_URLs = _WHISPER_TRANSCRIPT_URL + "/{split}-transcription.txt"
+_WHISPER_TRANSCRIPT_URLs = _WHISPER_TRANSCRIPT_URL + "/{split}-transcription.csv"
 
 
 class PeoplesSpeechConfig(datasets.BuilderConfig):
@@ -222,13 +222,11 @@ class PeoplesSpeech(datasets.GeneratorBasedBuilder):
                         "duration_ms": duration
                     }
 
-        whisper_transcripts = []
-
+        whisper_transcriptions = dict()
         with open(whisper_transcript, encoding="utf-8") as f:
-            for row in f:
-                whisper_transcripts.append(row.rstrip("\n"))
-
-        idx = 0
+            reader = csv.DictReader(f, delimiter=",", quoting=csv.QUOTE_NONE)
+            for line in tqdm(reader, desc="Reading transcriptions..."):
+                whisper_transcriptions[line["file_id"]] = line["whisper_transcript"]
 
 
         for local_extracted_archive_path, archive in zip(local_extracted_archive_paths, archives):
@@ -244,6 +242,5 @@ class PeoplesSpeech(datasets.GeneratorBasedBuilder):
                     "audio": {"path": path, "bytes": audio_file.read()},
                     "text": meta[audio_filename]["text"],
                     "duration_ms": meta[audio_filename]["duration_ms"],
-                    "whisper_transcript": whisper_transcripts[idx],
+                    "whisper_transcript": whisper_transcriptions.get(audio_filename, None),
                 }
-                idx += 1