This recipe contains the following corpora from LDC:

Audio:
  Gale phase 2/3/4
  LDC2013S08
  LDC2013S04
  LDC2014S09
  LDC2015S06
  LDC2015S13
  LDC2016S03
  LDC2017S25

  TDT 2/3/4
  LDC2001S93
  LDC2001S95
  LDC2005S11

Text:
  Gale phase 2/3/4
  LDC2013T20
  LDC2013T08
  LDC2014T28
  LDC2015T09
  LDC2015T25
  LDC2016T12
  LDC2017T18

  TDT 2/3/4
  LDC2001T57
  LDC2001T58
  LDC2005T16
  Besides, it uses Gigga word, simplified Mandarin for LM training and expanding dictionary:
  Gigga word (xin:simplified, cna:traditional. Use only xin)
  LDC2003T09
