Datasets:

cdminix
/

libriheavy

License:

Dataset card Files Files and versions Community

cdminix commited on Apr 15, 2024

Commit

ccc801d

·

verified ·

1 Parent(s): 7ebf1cd

Update libriheavy.py

Files changed (1) hide show

libriheavy.py +1 -1

libriheavy.py CHANGED Viewed

@@ -190,7 +190,7 @@ class Libriheavy(datasets.GeneratorBasedBuilder):
                         "speaker_name": chunk["speaker_name"],
                         "speaker_vec": npz_item["d_vector"][0],
                         "audio": chunk["audio"],
-                        "text": " ".join([segment[2] for segment in utterance["word_segments"] if "<" not in segment[2]])
                         "word_segments": [
                             {"start": segment[0], "end": segment[1], "word": segment[2]} for segment in utterance["word_segments"]
                         ],

                         "speaker_name": chunk["speaker_name"],
                         "speaker_vec": npz_item["d_vector"][0],
                         "audio": chunk["audio"],
+                        "text": " ".join([segment[2] for segment in utterance["word_segments"] if "<" not in segment[2]]),
                         "word_segments": [
                             {"start": segment[0], "end": segment[1], "word": segment[2]} for segment in utterance["word_segments"]
                         ],