“2013-04-19”版本间的差异

来自cslt Wiki
跳转至: 导航搜索
第25行: 第25行:
 
|speedup||26.8||27.88||26.81||22.82
 
|speedup||26.8||27.88||26.81||22.82
 
|-
 
|-
}
+
|}
 
:*Tencent baseline is with 700h online data+ 700h 863 data, HLDA+MPE, 88k lexicon
 
:*Tencent baseline is with 700h online data+ 700h 863 data, HLDA+MPE, 88k lexicon
 
:*Our results are with 400 hour AM, 88k LM. ML+bMMI
 
:*Our results are with 400 hour AM, 88k LM. ML+bMMI

2013年4月25日 (四) 04:22的版本

Data sharing

  • AM/lexicon/LM are shared.
  • LM count files are still in transfering.

DNN progress

400 hour DNN training

Test Set Tencent Baseline bMMI fMMI BN Hybrid
1900 8.4 7.65 7.35 6.57
2044 22.4 24.44 24.03 21.77
online1 35.6 34.66 34.33 31.44
online2 29.6 27.23 26.80 24.10
map 24.5 27.54 27.69 23.79
notepad 16 19.81 21.75 15.81
general 36 38.52 38.90 33.61
speedup 26.8 27.88 26.81 22.82
  • Tencent baseline is with 700h online data+ 700h 863 data, HLDA+MPE, 88k lexicon
  • Our results are with 400 hour AM, 88k LM. ML+bMMI

Tencent test result

AM: 70h training data(2 day, 15 machines, 10 threads)
LM: 88k LM
Test case: general
gmmi-bmmi: 38.7%
dnn-1: 28% 11 frame window, phone-based tree
dnn-2: 34% 9 frame window, state-based tree


GPU & CPU merge

Invesigate the possibility to merge GPU and CPU code. Try to find out an easier way. (1 week)

L-1 sparse initial training

Start to investigating.

Kaldi/HTK merge

  • HTK2Kaldi: the tool with Kaldi does not work.
  • Kaldi2HTK: done with implementation. Testing?

Embedded progress

  • Some large performance (speed) degradation with the embedded platform(1/60).
  • Planning for sparse DNN.
  • QA LM training, still failed. Mengyuan need more work on this.