You need to agree to share your contact information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this dataset content.

Stage 3 資料集組裝與切分統計報告

📊 全語料總覽 (Global Summary)

  • 有效語言總數: 6 個語言
  • 總語音樣本數: 18,121 筆
  • 總音訊總時長: 32.20 小時 (115,920.0 秒)
  • 切分策略: 確定性分層抽樣 (stratify_by=speaker, ratio=0.0%)
  • 排除重複音訊 (Deduplicated Audio): 3 筆 (0.00 小時, 文字不一致: 2 筆)
  • Train 集規模: 18,121 筆 (32.20 小時)
  • Eval 集規模: 0 筆 (0.00 小時, 佔比: 0.0%)

📋 各語言分切明細 (By Language)

語言別 (Language) ISO 代碼 Train 筆數 Train 時長 (hr) Eval 筆數 Eval 時長 (hr) 總筆數 總時長 (hr) Eval 佔比
南王卑南語 pyu-x-pym 1,973 2.19 0 0.00 1,973 2.19 0.0%
噶瑪蘭語 ckv 3,162 2.15 0 0.00 3,162 2.15 0.0%
太魯閣語 trv-x-truku 5,549 13.77 0 0.00 5,549 13.77 0.0%
建和卑南語 pyu-x-ksvk 2,507 4.13 0 0.00 2,507 4.13 0.0%
知本卑南語 pyu-x-ktrp 2,428 2.03 0 0.00 2,428 2.03 0.0%
西群卑南語 pyu-x-mkzy 2,502 7.93 0 0.00 2,502 7.93 0.0%
總計 (Total) - 18,121 32.20 0 0.00 18,121 32.20 0.0%
Downloads last month
46