From 9163ab1836c412d095f3d48d9271c2519b74d540 Mon Sep 17 00:00:00 2001 From: Vincent Quenneville-Belair Date: Fri, 25 Oct 2019 16:19:04 -0400 Subject: [PATCH 1/5] dataset path. --- test/test_datasets.py | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/test/test_datasets.py b/test/test_datasets.py index 4d8cacbb4a..8f35a27b56 100644 --- a/test/test_datasets.py +++ b/test/test_datasets.py @@ -7,9 +7,12 @@ from torchaudio.datasets.vctk import VCTK from torchaudio.datasets.yesno import YESNO +import common_utils + class TestDatasets(unittest.TestCase): - path = "assets" + test_dirpath, test_dir = common_utils.create_temp_assets_dir() + path = os.path.join(test_dirpath, "assets") def test_yesno(self): data = YESNO(self.path, return_dict=True) From 4c557597af15a1dfc06e701c56131ba472897d8f Mon Sep 17 00:00:00 2001 From: Vincent Quenneville-Belair Date: Fri, 25 Oct 2019 16:58:11 -0400 Subject: [PATCH 2/5] more explicit error. --- torchaudio/datasets/librispeech.py | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/torchaudio/datasets/librispeech.py b/torchaudio/datasets/librispeech.py index 59efe3fbc3..1ba4ccc9d8 100644 --- a/torchaudio/datasets/librispeech.py +++ b/torchaudio/datasets/librispeech.py @@ -1,7 +1,8 @@ import os -import torchaudio from torch.utils.data import Dataset + +import torchaudio from torchaudio.datasets.utils import ( download_url, extract_archive, @@ -32,7 +33,7 @@ def load_librispeech_item(fileid, path, ext_audio, ext_txt): break else: # Translation not found - raise ValueError + raise FileNotFoundError("Translation not found for " + file_audio) return { "speaker_id": speaker, From 75fa9dc986b98230799e695bd56256f3cf4f654d Mon Sep 17 00:00:00 2001 From: Vincent Quenneville-Belair Date: Fri, 25 Oct 2019 17:07:57 -0400 Subject: [PATCH 3/5] update txt file. --- .../LibriSpeech/dev-clean/1272/128104/1272-128104.trans.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/test/assets/LibriSpeech/dev-clean/1272/128104/1272-128104.trans.txt b/test/assets/LibriSpeech/dev-clean/1272/128104/1272-128104.trans.txt index bd06b7c4e2..57bf5c5086 100644 --- a/test/assets/LibriSpeech/dev-clean/1272/128104/1272-128104.trans.txt +++ b/test/assets/LibriSpeech/dev-clean/1272/128104/1272-128104.trans.txt @@ -1 +1 @@ -1272-128104-0000 THIS IS A TEST. +1272-128104-0000 MISTER QUILTER From 489ece0bbab41bd4970c5f27de63fa0493e2448b Mon Sep 17 00:00:00 2001 From: Vincent Quenneville-Belair Date: Fri, 25 Oct 2019 17:21:06 -0400 Subject: [PATCH 4/5] fix fileid_audio. --- torchaudio/datasets/librispeech.py | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/torchaudio/datasets/librispeech.py b/torchaudio/datasets/librispeech.py index 1ba4ccc9d8..2aab2117ed 100644 --- a/torchaudio/datasets/librispeech.py +++ b/torchaudio/datasets/librispeech.py @@ -20,7 +20,9 @@ def load_librispeech_item(fileid, path, ext_audio, ext_txt): file_text = speaker + "-" + chapter + ext_txt file_text = os.path.join(path, speaker, chapter, file_text) - file_audio = speaker + "-" + chapter + "-" + utterance + ext_audio + + fileid_audio = speaker + "-" + chapter + "-" + utterance + file_audio = file_audio + ext_audio file_audio = os.path.join(path, speaker, chapter, file_audio) # Load audio @@ -29,7 +31,7 @@ def load_librispeech_item(fileid, path, ext_audio, ext_txt): # Load text for line in open(file_text): fileid_text, content = line.strip().split(" ", 1) - if file_audio == fileid_text: + if fileid_audio == fileid_text: break else: # Translation not found From 515d481ab6da4a4d97536ac310335a91681480cc Mon Sep 17 00:00:00 2001 From: Vincent Quenneville-Belair Date: Fri, 25 Oct 2019 17:47:30 -0400 Subject: [PATCH 5/5] typo. --- torchaudio/datasets/librispeech.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/torchaudio/datasets/librispeech.py b/torchaudio/datasets/librispeech.py index 2aab2117ed..e9919ba1a3 100644 --- a/torchaudio/datasets/librispeech.py +++ b/torchaudio/datasets/librispeech.py @@ -22,7 +22,7 @@ def load_librispeech_item(fileid, path, ext_audio, ext_txt): file_text = os.path.join(path, speaker, chapter, file_text) fileid_audio = speaker + "-" + chapter + "-" + utterance - file_audio = file_audio + ext_audio + file_audio = fileid_audio + ext_audio file_audio = os.path.join(path, speaker, chapter, file_audio) # Load audio @@ -35,7 +35,7 @@ def load_librispeech_item(fileid, path, ext_audio, ext_txt): break else: # Translation not found - raise FileNotFoundError("Translation not found for " + file_audio) + raise FileNotFoundError("Translation not found for " + fileid_audio) return { "speaker_id": speaker,