“RNN test”版本间的差异
来自cslt Wiki
（→wsj_data） |
（→Test） |
||
| 第10行： | 第10行： | ||
[[process dict and data]] | [[process dict and data]] | ||
==Test== | ==Test== | ||
| − | [[ | + | [[wsj_data]] |
===RNNLM Rescore=== | ===RNNLM Rescore=== | ||
*Acoustic Model | *Acoustic Model | ||
2014年9月28日 (日) 11:23的版本
tool
- LTSM/RNN training, GPU&deep supported [1]
- RNNLM: RNN LM toolkit [2]
- RWTHLM: RNN LTSM toolkit [3]
- nplm: NN LM, large scale data [4]
- RNN toolkit from microsoft [5]
paper
Steps
Test
RNNLM Rescore
- Acoustic Model
- location: /nfs/disk/work/users/zhangzy/work/train_wsj_eng_new/data/train_si284
- test set
- location: /nfs/disk/work/users/zhangzy/work/train_wsj_eng_new/dt/test_eval92
- decode: /nfs/disk/work/users/zhangzy/work/train_wsj_eng_new/exp/tri4b_dnn_org/decode_eval92_tri4b_dnn_org
- Result
- lm:4.16%,rnnlm:3.47%
chinese data
prepare data
- now data
- gigaword: /work2/xingchao/corpus/Chinese_corpus/gigaword
- bing parallel corpus:/nfs/disk/work/users/xingchao/bing_dict
- baidu:
- sougou:
- using data
- sample gigword about 344M
- dict:tencent11w
- train set
| Parameters | hidden | class | direct | bbt | bptt_block | threads | direct-order | rand_seed | nwords | time(min) |
|---|---|---|---|---|---|---|---|---|---|---|
| set1 | 320 | 300 | 2000 | 2 | 20 | 1 | 4 | 1 | 10000 | 3380(56h) |
