find_alignment: word timings for textTokens within one 30 s window (mel [1, 3000, n_mels] model dtype, numFrames = content mel frames).
textTokens
mel
numFrames
find_alignment: word timings for
textTokenswithin one 30 s window (mel[1, 3000, n_mels] model dtype,numFrames= content mel frames).