app

scripts/runSQ.py

@@ -89,7 +89,8 @@ def align_human(meta,align_dir,speech_dir,model_path):
 
         caligner = aligner(model_path,model_word_sep,model_blank_tk)
         for rec in no_align:
-            wav_path = f'{speech_dir}{rec[1]}/{rec[2]}'
+            #wav_path = f'{speech_dir}{rec[1]}/{rec[2]}'
+            wav_path = f'{speech_dir}{rec[2]}'
             word_aln = caligner(wav16m(wav_path),rec[4],is_normed=True)
             apath = align_dir + rec[2].replace('.wav','.tsv')
             word_aln = [[str(x) for x in l] for l in word_aln]