Dataset Viewer
Auto-converted to Parquet Duplicate
id
stringlengths
28
32
speaker
stringclasses
12 values
text_id
stringclasses
131 values
epoch
int64
8
8
text
stringclasses
105 values
gold
stringclasses
111 values
audio
stringlengths
39
43
sr
int64
24k
24k
pred_dur
listlengths
163
404
align
dict
word_spans
dict
ep8_spk00_f_teen_bright_ls-0001
spk00_f_teen_bright
ls-0001
8
ช่วงนี้งานเยอะมากทั้งประชุมทั้งรายงานแทบไม่ได้พักเลย แต่สิ่งที่หนักที่สุดคือโปรเจกต์ใหม่ที่หัวหน้าเพิ่งมอบหมายให้เมื่อเช้าและย้ำว่ามี deadline
ช่วงนี้|งานเยอะมาก|ทั้งประชุม|ทั้งรายงาน|แทบไม่ได้พักเลย| แต่สิ่งที่หนักที่สุด|คือ|โปรเจกต์ใหม่|ที่หัวหน้า|เพิ่งมอบหมายให้เมื่อเช้า|และย้ำว่ามี deadline
audio/ep8_spk00_f_teen_bright_ls-0001.flac
24,000
[ 9, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 2, 1, 2, 2, 2, 2, 2, 3, 3, 1, 2, 2, 2, 1, 1, 1, 2, 1, 2, 1, 1, 1, 1, 1, 1, 2, 2, 2, 1, 1, 1, 1, 2, 2, 3, 1, 1, 1, 1, 2, 2, 1, 1, 1, 1, 2, 1, 2, 1, 1, 1, 1, 1...
{ "text_norm": "ช่วงนี้งานเยอะมากทั้งประชุมทั้งรายงานแทบไม่ได้พักเลย แต่สิ่งที่หนักที่สุดคือโปรเจกต์ใหม่ที่หัวหน้าเพิ่งมอบหมายให้เมื่อเช้าและย้ำว่ามี deadline", "phonemes": "cʰuːaŋ↘niː↗ŋaːn→jɤ↗maːk↘tʰaŋ↗pra˩cʰum→tʰaŋ↗raːj→ŋaːn→tʰɛːp↘maj↘daːj↘pʰak↗lɤːj→ tɛː˩siŋ˩tʰiː↘nak˩tʰiː↘sut˩kʰɯː→proː→ceːk˩maj˩tʰiː↘huːa↓naː↘pʰɤː...
{ "tltk": [ [ 0, 7, "ช่วงนี้" ], [ 7, 10, "งาน" ], [ 10, 14, "เยอะ" ], [ 14, 17, "มาก" ], [ 17, 21, "ทั้ง" ], [ 21, 27, "ประชุม" ], [ 27, 31, ...
ep8_spk00_f_teen_bright_ls-0002
spk00_f_teen_bright
ls-0002
8
Grab ที่เรียกเมื่อกี้แจ้งว่าอีกสิบห้านาทีถึง แต่ตอนนี้ผ่านมาครึ่งชั่วโมงแล้วคนขับยังวนอยู่แถวปากซอยหาทางเข้ามาไม่เจอสักที
Grab ที่เรียกเมื่อกี้แจ้งว่าอีกสิบห้านาทีถึง แต่ตอนนี้ผ่านมาครึ่งชั่วโมงแล้วคนขับยังวนอยู่แถวปากซอยหาทางเข้ามา|ไม่เจอสักที
audio/ep8_spk00_f_teen_bright_ls-0002.flac
24,000
[ 9, 1, 2, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 2, 2, 1, 1, 1, 1, 2, 1, 2, 2, 2, 1, 3, 1, 2, 2, 2, 2, 1, 1, 1, 1, 1, 1, 2, 1, 1, 2, 1, 2, 2, 2, 1, 1, 2, 1, 1, 1, 1, 2, 1, 2, 1, 1, 2, 2, 5, 3...
{ "text_norm": "Grab ที่เรียกเมื่อกี้แจ้งว่าอีกสิบห้านาทีถึง แต่ตอนนี้ผ่านมาครึ่งชั่วโมงแล้วคนขับยังวนอยู่แถวปากซอยหาทางเข้ามาไม่เจอสักที", "phonemes": "krɛp˩ tʰiː↘riːak↘mɯːa↘kiː↘cɛːŋ↘waː↘ʔiːk˩sip˩haː↘naː→tʰiː→tʰɯŋ↓ tɛː˩tɔːn→niː↗pʰaːn˩maː→kʰrɯŋ↘cʰuːa↘moːŋ→lɛːw↗kʰon→kʰap˩jaŋ→won→juː˩tʰɛːw↓paːk˩sɔːj→haː↓tʰaːŋ→kʰaw↘ma...
{ "tltk": [ [ 0, 4, "Grab" ], [ 5, 8, "ที่" ], [ 8, 13, "เรียก" ], [ 13, 21, "เมื่อกี้" ], [ 21, 25, "แจ้ง" ], [ 25, 28, "ว่า" ], [ 28, 31, ...
ep8_spk00_f_teen_bright_ls-0003
spk00_f_teen_bright
ls-0003
8
เมื่อวานได้มีโอกาสไปฟังบรรยายพิเศษของอาจารย์เกียรติศักดิ์ที่มหาวิทยาลัยซึ่งท่านเล่าเรื่องประวัติศาสตร์ไทยได้สนุกและเข้าใจง่ายมากจนเวลาสองชั่วโมงผ่านไปอย่างรวดเร็ว
เมื่อวาน|ได้มีโอกาส|ไปฟังบรรยายพิเศษ|ของอาจารย์เกียรติศักดิ์|ที่มหาวิทยาลัย|ซึ่งท่าน|เล่าเรื่องประวัติศาสตร์ไทย|ได้สนุก|และเข้าใจง่ายมาก|จน|เวลาสองชั่วโมง|ผ่านไป|อย่างรวดเร็ว
audio/ep8_spk00_f_teen_bright_ls-0003.flac
24,000
[ 10, 1, 2, 1, 1, 1, 1, 3, 2, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 2, 1, 3, 2, 2, 1, 1, 1, 1, 2, 2, 2, 2, 2, 1, 1, 1, 2, 1, 2, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, ...
{ "text_norm": "เมื่อวานได้มีโอกาสไปฟังบรรยายพิเศษของอาจารย์เกียรติศักดิ์ที่มหาวิทยาลัยซึ่งท่านเล่าเรื่องประวัติศาสตร์ไทยได้สนุกและเข้าใจง่ายมากจนเวลาสองชั่วโมงผ่านไปอย่างรวดเร็ว", "phonemes": "mɯːa↘waːn→daːj↘miː→ʔoː→kaːt˩paj→faŋ→ban→jaːj→pʰi↗seːt˩kʰɔːŋ↓ʔaː→caːn→kiːat˩ti˩sak˩tʰiː↘ma↗haː↓wit↗tʰa↗jaː→laj→sɯŋ↘tʰan↘law...
{ "tltk": [ [ 0, 8, "เมื่อวาน" ], [ 8, 11, "ได้" ], [ 11, 13, "มี" ], [ 13, 18, "โอกาส" ], [ 18, 20, "ไป" ], [ 20, 23, "ฟัง" ], [ 23, 29, "บ...
ep8_spk00_f_teen_bright_ls-0004
spk00_f_teen_bright
ls-0004
8
หลังจากที่ทีมงานช่วยกันปรับบทและถ่ายทำใหม่อยู่หลายรอบผลตอบรับจากผู้ชมก็ดีขึ้นเรื่อยเรื่อยโดยเฉพาะหลังจากที่ผู้กำกับได้อ่านความคิดเห็นและนำมาปรับตามซึ่งเราเรียกกระบวนการแบบนี้ว่า feedback
หลังจากที่|ทีมงาน|ช่วยกัน|ปรับบท|และถ่ายทำใหม่|อยู่หลายรอบ|ผลตอบรับ|จากผู้ชม|ก็ดีขึ้นเรื่อยเรื่อย|โดยเฉพาะ|หลังจากที่|ผู้กำกับ|ได้อ่าน|ความคิดเห็น|และนำมาปรับตาม|ซึ่งเราเรียก|กระบวนการแบบนี้|ว่า feedback
audio/ep8_spk00_f_teen_bright_ls-0004.flac
24,000
[ 10, 2, 2, 2, 2, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 2, 2, 1, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 2, 2, 3, 1, 1, 2, 2, 1, 1, 2, 4, 2, 1, 2, 2, 1, 2, 2, 1, 1, 1, 1, 1, 1, 2, 2, 1, 2, 2, ...
{ "text_norm": "หลังจากที่ทีมงานช่วยกันปรับบทและถ่ายทำใหม่อยู่หลายรอบผลตอบรับจากผู้ชมก็ดีขึ้นเรื่อยเรื่อยโดยเฉพาะหลังจากที่ผู้กำกับได้อ่านความคิดเห็นและนำมาปรับตามซึ่งเราเรียกกระบวนการแบบนี้ว่า feedback", "phonemes": "laŋ↓caːk˩tʰiː↘tʰiːm→ŋaːn→cʰuːaj↘kan→prap˩bot˩lɛ↗tʰaːj˩tʰam→maj˩juː˩laːj↓rɔːp↘pʰon↓tɔːp˩rap↗caːk˩pʰ...
{ "tltk": [ [ 0, 7, "หลังจาก" ], [ 7, 10, "ที่" ], [ 10, 16, "ทีมงาน" ], [ 16, 20, "ช่วย" ], [ 20, 23, "กัน" ], [ 23, 27, "ปรับ" ], [ 27, 29, ...
ep8_spk00_f_teen_bright_ls-0005
spk00_f_teen_bright
ls-0005
8
คุณยายเล่าให้ฟังว่าสมัยก่อนเวลามีงานบุญที่วัดชาวบ้านจะช่วยกันทำกระยาสารทแจกจ่ายกันทั้งหมู่บ้านเป็นภาพที่อบอุ่นและหาดูได้ยากในปัจจุบัน
คุณยายเล่าให้ฟังว่าสมัยก่อนเวลามีงานบุญที่วัด|ชาวบ้าน|จะช่วยกันทำกระยาสารท|แจกจ่ายกันทั้งหมู่บ้าน|เป็นภาพที่อบอุ่น|และหาดูได้ยาก|ในปัจจุบัน
audio/ep8_spk00_f_teen_bright_ls-0005.flac
24,000
[ 9, 1, 1, 1, 2, 2, 2, 3, 2, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 1, 2, 2, 2, 1, 1, 2, 2, 3, 1, 3, 1, 1, 2, 1, 1, 1, 1, 1, 3, 3, 3, 1, 2, 1, 2, 1, 2, 2, 1, 2, 1, 2, 2, 2, 1, 1, 1, 1, 1, 1, 2, 1, 2...
{ "text_norm": "คุณยายเล่าให้ฟังว่าสมัยก่อนเวลามีงานบุญที่วัดชาวบ้านจะช่วยกันทำกระยาสารทแจกจ่ายกันทั้งหมู่บ้านเป็นภาพที่อบอุ่นและหาดูได้ยากในปัจจุบัน", "phonemes": "kʰun→jaː→haj↘faŋ→waː↘sa˩maj↓kɔːn˩weː→laː→miː→ŋaːn→bun→tʰiː↘wat↗cʰaːw→baːn↘ca˩cʰuːaj↘kan→tʰam→kra˩jaː→saːt˩cɛːk˩caːj˩kan→tʰaŋ↗muː˩baːn↘pen→pʰaːp↘tʰiː↘ʔo...
{ "tltk": [ [ 0, 3, "คุณ" ], [ 3, 5, "ยา" ], [ 5, 10, "ยเล่า" ], [ 10, 13, "ให้" ], [ 13, 16, "ฟัง" ], [ 16, 19, "ว่า" ], [ 19, 27, "สมัยก่อ...
ep8_spk00_f_teen_bright_ls-0006
spk00_f_teen_bright
ls-0006
8
ปรารถนาสูงสุดของพ่อแม่ทุกคนคงไม่ใช่เรื่องเงินทองหรือชื่อเสียงแต่เป็นการได้เห็นลูกเติบโตขึ้นเป็นคนดีมีความสุขและยืนได้ด้วยลำแข้งของตัวเอง
ปรารถนาสูงสุด|ของพ่อแม่ทุกคน|คงไม่ใช่เรื่องเงินทอง|หรือชื่อเสียง|แต่เป็นการได้เห็นลูก|เติบโตขึ้น|เป็นคนดี|มีความสุข|และยืนได้|ด้วยลำแข้งของตัวเอง
audio/ep8_spk00_f_teen_bright_ls-0006.flac
24,000
[ 9, 1, 1, 3, 1, 1, 1, 1, 1, 1, 1, 1, 3, 2, 2, 2, 2, 1, 2, 2, 2, 2, 2, 2, 1, 1, 1, 1, 1, 1, 1, 2, 2, 1, 2, 1, 2, 2, 1, 1, 1, 1, 2, 1, 1, 1, 2, 2, 2, 1, 1, 2, 2, 2, 1, 1, 1, 1, 1, 1, 2, 2, 2, 1...
{ "text_norm": "ปรารถนาสูงสุดของพ่อแม่ทุกคนคงไม่ใช่เรื่องเงินทองหรือชื่อเสียงแต่เป็นการได้เห็นลูกเติบโตขึ้นเป็นคนดีมีความสุขและยืนได้ด้วยลำแข้งของตัวเอง", "phonemes": "praːt˩tʰa˩naː↓suːŋ↓sut˩kʰɔːŋ↓pʰɔː↘mɛː↘tʰuk↗kʰon→kʰoŋ→maj↘cʰaj↘rɯːaŋ↘ŋɤn→tʰɔːŋ→rɯː↓cʰɯː↘siːaŋ↓tɛː˩pen→kaːn→daːj↘hen↓luːk↘tɤːp˩toː→kʰɯn↘pen→kʰon→diː→m...
{ "tltk": [ [ 0, 7, "ปรารถนา" ], [ 7, 13, "สูงสุด" ], [ 13, 16, "ของ" ], [ 16, 22, "พ่อแม่" ], [ 22, 25, "ทุก" ], [ 25, 27, "คน" ], [ 27, 29, ...
ep8_spk00_f_teen_bright_ls-0007
spk00_f_teen_bright
ls-0007
8
ทริปนี้เราวางแผนขับรถออกจากกรุงเทพตั้งแต่เช้ามืดแวะกินข้าวระหว่างทางแล้วมุ่งหน้าไปทางอีสานโดยมีเป้าหมายปลายทางคือปราสาทหินพนมรุ้งที่ตั้งอยู่ในจังหวัดบุรีรัมย์
ทริปนี้|เราวางแผนขับรถ|ออกจากกรุงเทพ|ตั้งแต่เช้ามืด|แวะกินข้าว|ระหว่างทาง|แล้วมุ่งหน้า|ไปทางอีสาน|โดยมี|เป้าหมายปลายทาง|คือ|ปราสาทหินพนมรุ้ง|ที่ตั้งอยู่|ในจังหวัดบุรีรัมย์
audio/ep8_spk00_f_teen_bright_ls-0007.flac
24,000
[ 9, 2, 2, 2, 2, 1, 2, 2, 1, 1, 2, 2, 2, 2, 2, 1, 1, 2, 1, 2, 2, 1, 1, 2, 1, 1, 2, 2, 1, 2, 2, 2, 1, 1, 2, 1, 2, 2, 1, 2, 1, 2, 1, 1, 1, 1, 2, 2, 1, 1, 2, 1, 2, 2, 1, 1, 2, 2, 1, 2, 1, 1, 1, 1...
{ "text_norm": "ทริปนี้เราวางแผนขับรถออกจากกรุงเทพตั้งแต่เช้ามืดแวะกินข้าวระหว่างทางแล้วมุ่งหน้าไปทางอีสานโดยมีเป้าหมายปลายทางคือปราสาทหินพนมรุ้งที่ตั้งอยู่ในจังหวัดบุรีรัมย์", "phonemes": "sip↗niː↗raw→waːŋ→pʰɛːn↓kʰap˩rot↗ʔɔːk˩caːk˩kruŋ→tʰeːp↘taŋ↘tɛː˩cʰaw↗mɯːt↘wɛ↗kin→kʰaːw↘ra↗waːŋ˩tʰaːŋ→lɛːw↗muŋ↘naː↘paj→tʰaːŋ→ʔiː→s...
{ "tltk": [ [ 0, 4, "ทริป" ], [ 4, 7, "นี้" ], [ 7, 10, "เรา" ], [ 10, 16, "วางแผน" ], [ 16, 21, "ขับรถ" ], [ 21, 24, "ออก" ], [ 24, 27, "จา...
ep8_spk00_f_teen_bright_ls-0008
spk00_f_teen_bright
ls-0008
8
ภาพยนตร์เรื่องนี้เล่าเรื่องความเชื่อของคนโบราณที่มองว่าครุฑเป็นสัตว์ในตำนานซึ่งเป็นพาหนะของพระนารายณ์และเป็นสัญลักษณ์ของอำนาจจึงถูกนำมาใช้เป็นตราแผ่นดินมาจนถึงทุกวันนี้
ภาพยนตร์เรื่องนี้|เล่าเรื่อง|ความเชื่อของคนโบราณ|ที่มองว่า|ครุฑเป็นสัตว์ในตำนาน|ซึ่งเป็นพาหนะของพระนารายณ์|และเป็นสัญลักษณ์ของอำนาจ|จึงถูกนำมาใช้|เป็นตราแผ่นดิน|มาจนถึงทุกวันนี้
audio/ep8_spk00_f_teen_bright_ls-0008.flac
24,000
[ 9, 1, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 1, 2, 1, 1, 2, 1, 1, 2, 2, 1, 1, 2, 2, 2, 1, 2, 1, 1, 1, 2, 1, 1, 1, 2, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 2, 1, 2, 1, 2, 1, 1, 1...
{ "text_norm": "ภาพยนตร์เรื่องนี้เล่าเรื่องความเชื่อของคนโบราณที่มองว่าครุฑเป็นสัตว์ในตำนานซึ่งเป็นพาหนะของพระนารายณ์และเป็นสัญลักษณ์ของอำนาจจึงถูกนำมาใช้เป็นตราแผ่นดินมาจนถึงทุกวันนี้", "phonemes": "pʰaːp↘pʰa↗jon→rɯːaŋ↘niː↗law↘rɯːaŋ↘kʰwaːm→cʰɯːa↘kʰɔːŋ↓kʰon→boː→raːn→tʰiː↘mɔːŋ→waː↘kʰrut↗pen→sat˩naj→tam→naːn→sɯŋ↘pen→...
{ "tltk": [ [ 0, 8, "ภาพยนตร์" ], [ 8, 14, "เรื่อง" ], [ 14, 17, "นี้" ], [ 17, 27, "เล่าเรื่อง" ], [ 27, 36, "ความเชื่อ" ], [ 36, 39, "ของ" ], [ 39, ...
ep8_spk00_f_teen_bright_ls-0009
spk00_f_teen_bright
ls-0009
8
หมออธิบายว่าถ้าปล่อยให้ร่างกายมีพยาธิสะสมอยู่นานโดยไม่รักษาก็อาจส่งผลต่อการดูดซึมสารอาหารและทำให้ผู้ป่วยอ่อนเพลียเรื้อรังได้
หมออธิบายว่าถ้าปล่อยให้ร่างกาย|มีพยาธิสะสมอยู่นาน|โดยไม่รักษา|ก็อาจส่งผล|ต่อการดูดซึมสารอาหาร|และทำให้ผู้ป่วย|อ่อนเพลียเรื้อรังได้
audio/ep8_spk00_f_teen_bright_ls-0009.flac
24,000
[ 10, 1, 2, 2, 1, 1, 1, 1, 1, 1, 2, 2, 1, 3, 2, 1, 2, 2, 2, 1, 1, 1, 2, 2, 1, 2, 1, 1, 2, 2, 1, 1, 1, 1, 2, 1, 1, 2, 1, 2, 2, 1, 3, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, ...
{ "text_norm": "หมออธิบายว่าถ้าปล่อยให้ร่างกายมีพยาธิสะสมอยู่นานโดยไม่รักษาก็อาจส่งผลต่อการดูดซึมสารอาหารและทำให้ผู้ป่วยอ่อนเพลียเรื้อรังได้", "phonemes": "mɔː↓ʔa˩tʰi↗baːj→waː↘tʰaː↘plɔːj˩haj↘raːŋ↘kaːj→miː→pʰa↗jaː→tʰi↗sa˩som↓juː˩naːn→doːj→maj↘rak↗saː↓kɔː↘ʔaːt˩soŋ˩pʰon↓tɔː˩kaːn→duːt˩sɯm→saːn↓ʔaː→haːn↓lɛ↗tʰam→haj↘pʰuː...
{ "tltk": [ [ 0, 3, "หมอ" ], [ 3, 9, "อธิบาย" ], [ 9, 12, "ว่า" ], [ 12, 15, "ถ้า" ], [ 15, 20, "ปล่อย" ], [ 20, 23, "ให้" ], [ 23, 30, "ร่า...
ep8_spk00_f_teen_bright_ls-0010
spk00_f_teen_bright
ls-0010
8
สุดสัปดาห์ที่ผ่านมาแทบไม่ได้ออกไปไหนเลยเพราะฝนตกทั้งวันเลยนอนดูซีรีส์อยู่บ้านยาวยาวหลายเรื่องติดกันในแอป Netflix
สุดสัปดาห์ที่ผ่านมาแทบไม่ได้ออกไปไหนเลยเพราะฝนตกทั้งวัน|เลยนอนดูซีรีส์อยู่บ้านยาวยาว|หลายเรื่องติดกันในแอป Netflix
audio/ep8_spk00_f_teen_bright_ls-0010.flac
24,000
[ 10, 2, 2, 2, 2, 2, 2, 2, 1, 1, 3, 1, 1, 1, 1, 1, 1, 2, 1, 2, 2, 1, 2, 2, 1, 2, 1, 1, 1, 2, 2, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 2, 1, 1, 2, 2, 2, 1, 2, 2, 2, 5, 3, 1, ...
{ "text_norm": "สุดสัปดาห์ที่ผ่านมาแทบไม่ได้ออกไปไหนเลยเพราะฝนตกทั้งวันเลยนอนดูซีรีส์อยู่บ้านยาวยาวหลายเรื่องติดกันในแอป Netflix", "phonemes": "sut˩sap˩daː→tʰiː↘pʰaːn˩maː→tʰɛːp↘maj↘daːj↘ʔɔːk˩paj→naj↓lɤːj→pʰrɔ↗fon↓tok˩tʰaŋ↗wan→lɤːj→nɔːn→duː→siː→riː→juː˩baːn↘jaːw→jaːw→laːj↓rɯːaŋ↘tit˩kan→naj→ʔɛːp˩ net↗flik↗", "units...
{ "tltk": [ [ 0, 10, "สุดสัปดาห์" ], [ 10, 19, "ที่ผ่านมา" ], [ 19, 22, "แทบ" ], [ 22, 28, "ไม่ได้" ], [ 28, 31, "ออก" ], [ 31, 33, "ไป" ], [ 33, ...
ep8_spk00_f_teen_bright_ls-0011
spk00_f_teen_bright
ls-0011
8
สองพันยี่สิบเอ็ด และถือมาจนวันที่ ยี่สิบเก้า ธันวาคม ของปี สองพันยี่สิบเอ็ด จะคิดเป็นผลตอบแทนมหาศาลถึง สี่สิบเก้าล้าน เปอร์เซ็นต์ (แน่นอนว่าคงไม่น่ามีใครที่ซื้อและขายได้พอดีขนาดนี้)
สองพันยี่สิบเอ็ด| และถือมา|จนวันที่ ยี่สิบเก้า| ธันวาคม| ของปี| สองพันยี่สิบเอ็ด| จะคิดเป็น|ผลตอบแทน|มหาศาล|ถึง| สี่สิบเก้าล้าน| เปอร์เซ็นต์| (แน่นอนว่าคงไม่น่ามีใคร|ที่ซื้อและขาย|ได้พอดี|ขนาดนี้)
audio/ep8_spk00_f_teen_bright_ls-0011.flac
24,000
[ 9, 2, 2, 1, 2, 2, 1, 1, 2, 2, 2, 2, 1, 1, 2, 1, 1, 2, 2, 1, 3, 4, 2, 2, 1, 2, 2, 1, 1, 2, 1, 2, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 1, 3, 2, 2...
{ "text_norm": "สองพันยี่สิบเอ็ด และถือมาจนวันที่ ยี่สิบเก้า ธันวาคม ของปี สองพันยี่สิบเอ็ด จะคิดเป็นผลตอบแทนมหาศาลถึง สี่สิบเก้าล้าน เปอร์เซ็นต์ (แน่นอนว่าคงไม่น่ามีใครที่ซื้อและขายได้พอดีขนาดนี้)", "phonemes": "sɔːŋ↓pʰan→jiː↘sip˩ʔet˩ lɛ↗tʰɯː↓maː→ca˩na↗wan→tʰiː↘ jiː↘sip˩kaw↘ tʰan→waː→kʰom→ kʰɔːŋ↓piː→ sɔːŋ↓pʰan→jiː...
{ "tltk": [ [ 0, 3, "สอง" ], [ 3, 6, "พัน" ], [ 6, 9, "ยี่" ], [ 9, 12, "สิบ" ], [ 12, 16, "เอ็ด" ], [ 17, 20, "และ" ], [ 20, 23, "ถือ" ...
ep8_spk00_f_teen_bright_ls-0012
spk00_f_teen_bright
ls-0012
8
กองทุนที่เน้นลงทุนในธุรกิจ Digital Health ที่เกี่ยวข้องกับบริการ Telehealth การใช้ระบบ Cloud ในการจัดเก็บข้อมูลผู้ป่วย รวมถึงบริษัทที่ผลิตเครื่องมือการตรวจโรค
กองทุน|ที่เน้นลงทุน|ในธุรกิจ| Digital Health| ที่เกี่ยวข้อง|กับบริการ| Telehealth| การใช้ระบบ| Cloud| ในการจัดเก็บ|ข้อมูลผู้ป่วย| รวมถึงบริษัท|ที่ผลิตเครื่องมือ|การตรวจโรค
audio/ep8_spk00_f_teen_bright_ls-0012.flac
24,000
[ 9, 1, 2, 1, 2, 1, 1, 1, 2, 2, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 2, 2, 2, 1, 1, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 3, 1, 3, 3, 2, 1, 1, 2, 2, 1, 1, 2, 1, 2, 3, 1, 2, 2, 2, 2, 2, 1, 1, 1...
{ "text_norm": "กองทุนที่เน้นลงทุนในธุรกิจ Digital Health ที่เกี่ยวข้องกับบริการ Telehealth การใช้ระบบ Cloud ในการจัดเก็บข้อมูลผู้ป่วย รวมถึงบริษัทที่ผลิตเครื่องมือการตรวจโรค", "phonemes": "kɔːŋ→tʰun→tʰiː↘neːn↗loŋ→tʰun→naj→tʰu↗ra↗kit˩ di˩ci˩tʰan→ heːn→ tʰiː↘kiːaw˩kʰɔːŋ↘kap˩bɔː→ri↗kaːn→ tʰeː→leː→heːn→ kaːn→cʰaj↗ra↗b...
{ "tltk": [ [ 0, 6, "กองทุน" ], [ 6, 9, "ที่" ], [ 9, 13, "เน้น" ], [ 13, 18, "ลงทุน" ], [ 18, 20, "ใน" ], [ 20, 26, "ธุรกิจ" ], [ 27, 34, "...
ep8_spk00_f_teen_bright_ls-0013
spk00_f_teen_bright
ls-0013
8
กองทุนที่เน้นลงทุนในธุรกิจ Digital Health ที่เกี่ยวข้องกับบริการ Telehealth การใช้ระบบ Cloud ในการจัดเก็บข้อมูลผู้ป่วย รวมถึงบริษัทที่ผลิตเครื่องมือการตรวจโรค
กองทุน|ที่เน้นลงทุน|ในธุรกิจ| Digital Health| ที่เกี่ยวข้อง|กับบริการ| Telehealth| การใช้ระบบ| Cloud| ในการจัดเก็บ|ข้อมูลผู้ป่วย| รวมถึงบริษัท|ที่ผลิต|เครื่องมือการตรวจโรค
audio/ep8_spk00_f_teen_bright_ls-0013.flac
24,000
[ 9, 1, 2, 1, 2, 1, 1, 1, 2, 2, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 2, 2, 2, 1, 1, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 3, 1, 3, 3, 2, 1, 1, 2, 2, 1, 1, 2, 1, 2, 3, 1, 2, 2, 2, 2, 2, 1, 1, 1...
{ "text_norm": "กองทุนที่เน้นลงทุนในธุรกิจ Digital Health ที่เกี่ยวข้องกับบริการ Telehealth การใช้ระบบ Cloud ในการจัดเก็บข้อมูลผู้ป่วย รวมถึงบริษัทที่ผลิตเครื่องมือการตรวจโรค", "phonemes": "kɔːŋ→tʰun→tʰiː↘neːn↗loŋ→tʰun→naj→tʰu↗ra↗kit˩ di˩ci˩tʰan→ heːn→ tʰiː↘kiːaw˩kʰɔːŋ↘kap˩bɔː→ri↗kaːn→ tʰeː→leː→heːn→ kaːn→cʰaj↗ra↗b...
{ "tltk": [ [ 0, 6, "กองทุน" ], [ 6, 9, "ที่" ], [ 9, 13, "เน้น" ], [ 13, 18, "ลงทุน" ], [ 18, 20, "ใน" ], [ 20, 26, "ธุรกิจ" ], [ 27, 34, "...
ep8_spk00_f_teen_bright_ls-0014
spk00_f_teen_bright
ls-0014
8
กองทุนที่เน้นลงทุนในธุรกิจ Digital Health ที่เกี่ยวข้องกับบริการ Telehealth การใช้ระบบ Cloud ในการจัดเก็บข้อมูลผู้ป่วย รวมถึงบริษัทที่ผลิตเครื่องมือการตรวจโรค
กองทุน|ที่เน้นลงทุน|ในธุรกิจ| Digital Health| ที่เกี่ยวข้อง|กับบริการ| Telehealth| การใช้ระบบ| Cloud| ในการจัดเก็บ|ข้อมูลผู้ป่วย| รวมถึงบริษัท|ที่ผลิตเครื่องมือ|การตรวจโรค
audio/ep8_spk00_f_teen_bright_ls-0014.flac
24,000
[ 9, 1, 2, 1, 2, 1, 1, 1, 2, 2, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 2, 2, 2, 1, 1, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 3, 1, 3, 3, 2, 1, 1, 2, 2, 1, 1, 2, 1, 2, 3, 1, 2, 2, 2, 2, 2, 1, 1, 1...
{ "text_norm": "กองทุนที่เน้นลงทุนในธุรกิจ Digital Health ที่เกี่ยวข้องกับบริการ Telehealth การใช้ระบบ Cloud ในการจัดเก็บข้อมูลผู้ป่วย รวมถึงบริษัทที่ผลิตเครื่องมือการตรวจโรค", "phonemes": "kɔːŋ→tʰun→tʰiː↘neːn↗loŋ→tʰun→naj→tʰu↗ra↗kit˩ di˩ci˩tʰan→ heːn→ tʰiː↘kiːaw˩kʰɔːŋ↘kap˩bɔː→ri↗kaːn→ tʰeː→leː→heːn→ kaːn→cʰaj↗ra↗b...
{ "tltk": [ [ 0, 6, "กองทุน" ], [ 6, 9, "ที่" ], [ 9, 13, "เน้น" ], [ 13, 18, "ลงทุน" ], [ 18, 20, "ใน" ], [ 20, 26, "ธุรกิจ" ], [ 27, 34, "...
ep8_spk00_f_teen_bright_ls-0015
spk00_f_teen_bright
ls-0015
8
กองทุนที่เน้นลงทุนในธุรกิจ Digital Health ที่เกี่ยวข้องกับบริการ Telehealth การใช้ระบบ Cloud ในการจัดเก็บข้อมูลผู้ป่วย รวมถึงบริษัทที่ผลิตเครื่องมือการตรวจโรค
กองทุน|ที่เน้นลงทุน|ในธุรกิจ| Digital Health| ที่เกี่ยวข้อง|กับบริการ| Telehealth| การใช้ระบบ| Cloud| ในการจัดเก็บ|ข้อมูลผู้ป่วย| รวมถึงบริษัท|ที่ผลิต|เครื่องมือการตรวจโรค
audio/ep8_spk00_f_teen_bright_ls-0015.flac
24,000
[ 9, 1, 2, 1, 2, 1, 1, 1, 2, 2, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 2, 2, 2, 1, 1, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 3, 1, 3, 3, 2, 1, 1, 2, 2, 1, 1, 2, 1, 2, 3, 1, 2, 2, 2, 2, 2, 1, 1, 1...
{ "text_norm": "กองทุนที่เน้นลงทุนในธุรกิจ Digital Health ที่เกี่ยวข้องกับบริการ Telehealth การใช้ระบบ Cloud ในการจัดเก็บข้อมูลผู้ป่วย รวมถึงบริษัทที่ผลิตเครื่องมือการตรวจโรค", "phonemes": "kɔːŋ→tʰun→tʰiː↘neːn↗loŋ→tʰun→naj→tʰu↗ra↗kit˩ di˩ci˩tʰan→ heːn→ tʰiː↘kiːaw˩kʰɔːŋ↘kap˩bɔː→ri↗kaːn→ tʰeː→leː→heːn→ kaːn→cʰaj↗ra↗b...
{ "tltk": [ [ 0, 6, "กองทุน" ], [ 6, 9, "ที่" ], [ 9, 13, "เน้น" ], [ 13, 18, "ลงทุน" ], [ 18, 20, "ใน" ], [ 20, 26, "ธุรกิจ" ], [ 27, 34, "...
ep8_spk00_f_teen_bright_ls-0016
spk00_f_teen_bright
ls-0016
8
การปั๊มชีพจรเศรษฐกิจจากปัญหาเงินเฟ้อแบบหนักมือ Demand ฟื้นขึ้นมาอย่างรวดเร็ว แต่ Supply ตาม Demand ขึ้นมาไม่ทัน
การปั๊มชีพจรเศรษฐกิจ|จากปัญหาเงินเฟ้อ|แบบหนักมือ| Demand| ฟื้นขึ้นมา|อย่างรวดเร็ว| แต่ Supply| ตาม| Demand| ขึ้นมาไม่ทัน
audio/ep8_spk00_f_teen_bright_ls-0016.flac
24,000
[ 9, 1, 2, 1, 2, 2, 1, 2, 2, 2, 1, 1, 2, 2, 1, 1, 1, 1, 1, 2, 1, 2, 1, 1, 2, 2, 1, 1, 1, 1, 1, 1, 1, 2, 1, 2, 2, 2, 1, 1, 1, 2, 1, 1, 1, 1, 2, 2, 3, 2, 2, 1, 2, 2, 2, 2, 2, 2, 1, 1, 2, 1, 2, 2...
{ "text_norm": "การปั๊มชีพจรเศรษฐกิจจากปัญหาเงินเฟ้อแบบหนักมือ Demand ฟื้นขึ้นมาอย่างรวดเร็ว แต่ Supply ตาม Demand ขึ้นมาไม่ทัน", "phonemes": "kaːn→pam↗cʰiːp↘pʰa↗cɔːn→seːt˩tʰa˩kit˩caːk˩pan→haː↓ŋɤn→fɤː↗bɛːp˩nak˩mɯː→ diː→maːn→ fɯːn↗kʰɯn↘maː→jaːŋ˩ruːat↘rew→ tɛː˩ sap↗pʰlaːj→ taːm→ diː→maːn→ kʰɯn↘maː→maj↘tʰan→", "unit...
{ "tltk": [ [ 0, 3, "การ" ], [ 3, 7, "ปั๊ม" ], [ 7, 12, "ชีพจร" ], [ 12, 20, "เศรษฐกิจ" ], [ 20, 23, "จาก" ], [ 23, 28, "ปัญหา" ], [ 28, 36, ...
ep8_spk00_f_teen_bright_ls-0017
spk00_f_teen_bright
ls-0017
8
นักลงทุนมีแนวโน้มที่จะย้ายไปลงทุนในกลุ่ม Value Stock ที่มีผลตอบแทนจากเงินปันผลและค่อนข้างปลอดภัยกว่า หุ้น Value มีโอกาสปรับราคาขึ้น
นักลงทุน|มีแนวโน้ม|ที่จะ|ย้ายไปลงทุน|ในกลุ่ม| Value Stock |ที่มี|ผลตอบแทน|จากเงินปันผล|และค่อนข้างปลอดภัยกว่า| หุ้น Value |มีโอกาส|ปรับราคาขึ้น
audio/ep8_spk00_f_teen_bright_ls-0017.flac
24,000
[ 10, 1, 2, 2, 2, 1, 2, 2, 2, 1, 1, 2, 2, 2, 1, 2, 1, 1, 1, 2, 1, 2, 1, 1, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 3, 1, 1, 2, 1, 1, 1, 1, 1, 1, 2, 2, 1, 1, 2, 2, 1, 1, 1, 1, 3, 1, 1, 3, 3, 7, ...
{ "text_norm": "นักลงทุนมีแนวโน้มที่จะย้ายไปลงทุนในกลุ่ม Value Stock ที่มีผลตอบแทนจากเงินปันผลและค่อนข้างปลอดภัยกว่า หุ้น Value มีโอกาสปรับราคาขึ้น", "phonemes": "nak↗loŋ→tʰun→miː→nɛːw→noːm↗tʰiː↘ca˩jaːj↗paj→loŋ→tʰun→naj→klum˩ wɛː→luː→ sa˩tɔk˩ tʰiː↘miː→pʰon↓tɔːp˩tʰɛːn→caːk˩ŋɤn→pan→pʰon↓lɛ↗kʰɔːn↘kʰaːŋ↘plɔːt˩pʰaj→kwaː...
{ "tltk": [ [ 0, 8, "นักลงทุน" ], [ 8, 10, "มี" ], [ 10, 17, "แนวโน้ม" ], [ 17, 20, "ที่" ], [ 20, 22, "จะ" ], [ 22, 26, "ย้าย" ], [ 26, 28, ...
ep8_spk00_f_teen_bright_ls-0018
spk00_f_teen_bright
ls-0018
8
นักลงทุนมีแนวโน้มที่จะย้ายไปลงทุนในกลุ่ม Value Stock ที่มีผลตอบแทนจากเงินปันผลและค่อนข้างปลอดภัยกว่า หุ้น Value มีโอกาสปรับราคาขึ้น
นักลงทุน|มีแนวโน้ม|ที่จะ|ย้ายไปลงทุน|ในกลุ่ม| Value Stock |ที่มี|ผลตอบแทน|จากเงินปันผล|และค่อนข้างปลอดภัยกว่า| หุ้น Value |มีโอกาส|ปรับราคาขึ้น
audio/ep8_spk00_f_teen_bright_ls-0018.flac
24,000
[ 10, 1, 2, 2, 2, 1, 2, 2, 2, 1, 1, 2, 2, 2, 1, 2, 1, 1, 1, 2, 1, 2, 1, 1, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 3, 1, 1, 2, 1, 1, 1, 1, 1, 1, 2, 2, 1, 1, 2, 2, 1, 1, 1, 1, 3, 1, 1, 3, 3, 7, ...
{ "text_norm": "นักลงทุนมีแนวโน้มที่จะย้ายไปลงทุนในกลุ่ม Value Stock ที่มีผลตอบแทนจากเงินปันผลและค่อนข้างปลอดภัยกว่า หุ้น Value มีโอกาสปรับราคาขึ้น", "phonemes": "nak↗loŋ→tʰun→miː→nɛːw→noːm↗tʰiː↘ca˩jaːj↗paj→loŋ→tʰun→naj→klum˩ wɛː→luː→ sa˩tɔk˩ tʰiː↘miː→pʰon↓tɔːp˩tʰɛːn→caːk˩ŋɤn→pan→pʰon↓lɛ↗kʰɔːn↘kʰaːŋ↘plɔːt˩pʰaj→kwaː...
{ "tltk": [ [ 0, 8, "นักลงทุน" ], [ 8, 10, "มี" ], [ 10, 17, "แนวโน้ม" ], [ 17, 20, "ที่" ], [ 20, 22, "จะ" ], [ 22, 26, "ย้าย" ], [ 26, 28, ...
ep8_spk00_f_teen_bright_ls-0019
spk00_f_teen_bright
ls-0019
8
แม้รายได้จะสูงเมื่อเทียบอดีต แต่การเติบโตเทียบช่วงเดียวกันของปี สองพันยี่สิบเอ็ด เปอร์เซ็นต์ YoY เหลือเพียง ห้า เปอร์เซ็นต์ ก็คือแทบจะไม่โตแล้ว ซึ่งเกิดขึ้นจากทั้งกำลังการผลิตเต็ม
แม้รายได้จะสูง|เมื่อเทียบอดีต| แต่การเติบโต|เทียบช่วงเดียวกัน|ของปี สองพันยี่สิบเอ็ด| เปอร์เซ็นต์ YoY| เหลือเพียง ห้า เปอร์เซ็นต์| ก็คือ|แทบจะไม่โตแล้ว ซึ่งเกิดขึ้น|จากทั้ง|กำลังการผลิตเต็ม
audio/ep8_spk00_f_teen_bright_ls-0019.flac
24,000
[ 10, 1, 3, 2, 1, 1, 3, 1, 1, 1, 1, 2, 1, 1, 2, 1, 2, 2, 2, 2, 2, 2, 2, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 3, 4, 8, 3, 2, 1, 1, 1, 2, 1, 2, 1, 1, 2, 1, 2, 1, 2, 2, 1, 2, 2, 1, 1, ...
{ "text_norm": "แม้รายได้จะสูงเมื่อเทียบอดีต แต่การเติบโตเทียบช่วงเดียวกันของปี สองพันยี่สิบเอ็ด เปอร์เซ็นต์ YoY เหลือเพียง ห้า เปอร์เซ็นต์ ก็คือแทบจะไม่โตแล้ว ซึ่งเกิดขึ้นจากทั้งกำลังการผลิตเต็ม", "phonemes": "mɛː↗raːj→daːj↘ca˩suːŋ↓mɯːa↘tʰiːap↘ʔa˩diːt˩ tɛː˩kaːn→tɤːp˩toː→tʰiːap↘cʰuːaŋ↘diːaw→kan→kʰɔːŋ↓piː→ sɔːŋ↓pʰan...
{ "tltk": [ [ 0, 3, "แม้" ], [ 3, 9, "รายได้" ], [ 9, 11, "จะ" ], [ 11, 14, "สูง" ], [ 14, 19, "เมื่อ" ], [ 19, 24, "เทียบ" ], [ 24, 28, "อด...
End of preview. Expand in Data Studio

Thai Aligner Bench

🚧 Development in progress. This benchmark is under active development. The aligner set, the metrics and the asset bundle may all change without notice, and every number below should be read as provisional. If you need reproducibility, pin a commit SHA rather than tracking main.

Which forced aligner should score Thai TTS pauses? A self-contained benchmark: one Python file (aligner_bench.py) plus 1,572 clips of Thai TTS audio carrying frozen timing ground truth. No Thai NLP, no TTS model, no other code needed — just numpy soundfile torch torchaudio transformers.

The ground truth is unusual and is what makes this dataset useful: the audio was rendered by a TTS model whose duration predictor emitted an explicit frame count per token (pred_dur, 25 ms/frame, sum(pred_dur) * 600 == len(audio) exactly at 24 kHz). The token boundaries in the audio therefore are the pred_dur boundaries — no human annotation, no second aligner, no circularity. Any forced aligner can be measured against them directly.

Two questions are answered per aligner:

  1. Delay / timing accuracy — onset error percentiles, signed bias, jitter (spread with bias removed), duration error, span coverage.
  2. Downstream pause accuracy — silences are detected, attributed to a text juncture through the aligner's char timings, and classified ok / bad_juncture / intra_word / unaligned. Scoring the same clips through pred_dur gives the reference, so each aligner also reports agreement with ground truth.

Quick start

git clone https://huggingface.co/datasets/kunato/thai-aligner-bench-dev
cd thai-aligner-bench-dev
pip install numpy soundfile torch torchaudio transformers

python aligner_bench.py run --bundle . --aligner ctc --out out/ctc
python aligner_bench.py run --bundle . --aligner mms --out out/mms
python aligner_bench.py report out/ctc out/mms

run = align (GPU, writes spans.jsonl) + score (CPU, writes metrics.json). Run them separately to re-score without re-aligning. Pre-computed spans.jsonl for all four arms is in results/, so you can reproduce every number below on CPU alone:

cp -r results/ctc out/ctc && python aligner_bench.py score --bundle . --out out/ctc

Useful flags: --limit N / --speakers spk00,spk03 (fast subsets), --word-seg {tltk,newmm} (which frozen word segmentation attributes the pause), --quantize-ms 38.5 (round predicted onsets onto a coarser grid — tests whether an aligner's deficit is really just its frame rate), --device, --batch-size.

Aligners included

arm model note
ctc airesearch/wav2vec2-large-xlsr-53-th Thai char CTC, 50 fps — the default
mms torchaudio.pipelines.MMS_FA generic multilingual phone aligner; IPA mapped onto its roman labels
qwen Qwen/Qwen3-ASR-1.7B-hf + a Thai char timestamp head --ctc-head-repo; head architecture is inferred from the weights, so a new head loads with no code change

The qwen arm needs transformers>=5.13. Side-load rather than upgrade if pinned:

pip install --no-deps --target=/tmp/tf514 transformers==5.14.1
PYTHONPATH=/tmp/tf514 python aligner_bench.py run --bundle . --aligner qwen \
    --ctc-head-repo <head-repo> --out out/qwen

Add your own aligner

Write one function returning [(char_idx, t0_ms, t1_ms)] over item["text"], time-monotonic, and register it:

def mine_bundle(device, **_):
    return SimpleNamespace(..., tag="mine:v1", batched=False)

def mine_char_spans(wav, sr, item, b):
    ...   # item["_align"].phonemes / .units are there if you align on phones
    return spans

ALIGNERS["mine"] = (mine_bundle, mine_char_spans, None)

Everything downstream — silence detection, attribution, classification, aggregation, agreement — is shared, so arms stay comparable.

If you batch, pass an explicit length mask. This is not a style note. wav2vec2-large-xlsr-53-th ships return_attention_mask: false while being a feat_extract_norm='layer' model trained with masking. Running it unmasked scales each clip's whole char timeline by len/batch_max — drift that reached −1670 ms by the end of a 7.5 s clip, silently destroying every downstream number until it was found. The Qwen arm has the identical trap (features padded to 30 s). Both shipped paths build the mask from raw input lengths.

Results

1,572 clips (12 synthetic speakers × 131 long-sentence texts), --word-seg tltk, default thresholds, one RTX 3090. Produced by this exact file; raw artifacts in results/.

Timing vs pred_dur ground truth (ms; lower is better):

model fps onset p50 p90 >100 ms bias jitter dur p50 coverage s/clip
ctc xlsr-53-th 50 17.26 40.74 0.2% −15.91 11.96 15.21 0.401 0.209
mms MMS_FA 50 28.43 46.91 0.4% −28.39 9.38 14.52 0.709 0.154
qwen + ctc-head-v2 26 36.54 82.20 4.4% −33.62 24.09 28.44 0.669 0.331
qwen + ctc-head-cv-only 26 29.97 72.52 2.9% −26.59 22.34 27.72 0.731 0.332

Downstream pause metrics + agreement with ground truth:

model /clip precision intra_word PPER bad/min r(count) verdict per-spk r bad-junc Jaccard
pred_dur (truth) 2.291 0.8681 0.0878 0.2226 1.911
ctc 2.268 0.8634 0.1018 0.2290 1.959 0.9603 95.2% 0.9614 248/319 0.6685
mms 2.414 0.7718 0.2265 0.3321 3.484 0.9411 85.4% 0.7525 249/319 0.6587
qwen + v2 2.240 0.8606 0.1031 0.2284 1.975 0.9597 94.3% 0.9353 241/319 0.6359
qwen + cv-only 2.233 0.8390 0.1387 0.2665 2.273 0.9611 90.5% 0.9388 240/319 0.6107

ctc wins every downstream column. Three results worth more than the ranking:

  • Timing accuracy does not imply attribution accuracy. MMS has the tightest jitter of any arm (9.38 ms vs ctc's 11.96) and the best duration error — once its larger constant bias is removed it is the sharpest aligner here. It still loses 0.09 precision, drops per-speaker r to 0.75, and more than doubles intra_word. It aligns romanized Thai through a lossy map (length and tone dropped; ɛ/e, ɔ/o, ɯ/ɤ/u collapse), so it puts precisely-timed boundaries in the wrong places. Character identity beats milliseconds.
  • Below ~40 ms median onset error, timing stops mattering for this metric. qwen+v2 has 2× ctc's onset error and 2× its duration error, yet lands within 0.003 precision and 0.026 per-speaker r. Pauses are ≥80 ms, so a 36 ms error rarely flips which juncture a silence belongs to. Do not tune an aligner for this metric past ~40 ms — spend the effort on character/tokenizer fidelity instead.
  • The qwen heads' deficit is not their frame rate. Rounding ctc's own onsets onto qwen's 38.5 ms grid (--quantize-ms 38.5) moves it 17.26 → 19.0 ms p50 and 11.96 → 14.5 ms jitter. The grid costs ~1.7 ms against a ~19 ms gap. A bigger ConvTranspose upsample would not help either: it makes a finer output grid over the encoder's native 13 fps / 77 ms information floor. Beating ctc needs a finer-resolution encoder, not a bigger head.

Between the two qwen heads, v2 (GigaSpeech2-th) beats cv-only (Common Voice) downstream — 0.861 vs 0.839 precision, 94.3% vs 90.5% verdict agreement — despite cv-only's better raw onsets (29.97 vs 36.54). Same architecture, so this is a training-data effect, and it is the third instance here of better timing not surviving into better attribution.

Reading the numbers

  • Read precision next to /clip. The masks are precision-side only — they mark where a break is allowed, never where one is required — so an aligner (or a TTS system) that finds no pauses scores 1.000. Precision alone is not a ranking.
  • per-spk r is the real instrument test. It is whether the aligner ranks the 12 voices the way ground truth does. An aligner can hold aggregate precision and still be useless for comparing systems; this column catches it.
  • bias is correctable, jitter is not. Aligners report the left edge of a frame, so each carries a systematic half-frame early bias that a constant shift removes. Jitter is what decides usability.
  • Never compare CTC span widths across aligners. They allocate blank frames differently — coverage runs from 0.40 to 0.73 here. Measuring duration as a span's own width reverses the ranking. dur p50 therefore uses onset[i+1] − onset[i], the convention pred_dur itself uses.

Dataset structure

items.jsonl        one JSON per clip (1,572)
audio/<id>.flac    24 kHz mono, lossless
aligner_bench.py   the benchmark (single file)
results/<arm>/     metrics.json, align_meta.json, spans.jsonl
field meaning
id / speaker / text_id clip key; text_id groups the same text across the 12 voices
text the normalized text actually spoken — every char index, gold mark and word span below is in this coordinate space
gold pause mask; | marks a juncture where a break is allowed (hand-calibrated, precision-side only)
pred_dur duration-predictor frame counts, 25 ms/frame, one pad token at each end — the timing ground truth
align.phonemes IPA, byte-identical to what the model was fed
align.units syllable / English-word / punct atoms: [ph_a, ph_b) in phonemes, [ch_a, ch_b) in text
align.words word spans with surfaces
align.ph2unit phoneme char → unit index (−1 for spaces)
align.tok2ph model token index → phoneme char index
word_spans two independent word segmentations (tltk, newmm) so tokenizer choice can be varied with no Thai tokenizer installed

Limitations

  • The audio is vocoder output, not natural speech. It is out of domain for aligners trained on read speech. That is deliberate — it is the audio the metric has to work on — but results here are not a claim about natural speech, and an aligner that does well on real speech may not do well here.
  • pred_dur is ground truth for this TTS model's rendering, not a human phonetic transcription. It is exact about where the model placed each token; it does not adjudicate whether that placement was phonetically ideal.
  • Precision-side only. No "a break is required here" marks exist, so missing a break is unpenalized under pause_precision.
  • Synthetic speakers. 12 designed voices from one checkpoint, so speaker diversity is narrower than a natural-speech corpus.
  • Earlier internal runs of this comparison reported larger gaps between aligners. Two things changed: a coordinate bug specific to the MMS arm (its char indices came from the G2P's units, which index normalized text, while attribution used raw text — they differ on 900/1,572 clips), and everything now aligning on the text actually spoken. Both raise agreement with ground truth across the board. The ranking is unchanged; the margins were overstated.
Downloads last month
38