Loading the LM will be faster if you build a binary file.
Reading ../standard/nc-wmt11.en.srilm.gz
----5---10---15---20---25---30---35---40---45---50---55---60---65---70---75---80---85---90---95--100
****************************************************************************************************
  Example feature: Shape_S00000_T00000
dtrain
Parameters:
                       k 100
                       N 4
                       T 1
           learning rate 0.0001
                  margin 1
                  l1 reg 0
            decoder conf 'cdec.ini'
                   input 'work/shard.2.0'
                  output 'work/weights.2.2'
              weights in 'work/weights.1'
(1 dot per processed input)
Iteration #1 of 1.
 ... 3
WEIGHTS
              Glue = -0.1781
       WordPenalty = +0.262
     LanguageModel = +1.468
 LanguageModel_OOV = -0.4828
     PhraseModel_0 = -0.4867
     PhraseModel_1 = +0.3306
     PhraseModel_2 = +0.7159
     PhraseModel_3 = +0.01232
     PhraseModel_4 = -1.128
     PhraseModel_5 = +0.04098
     PhraseModel_6 = -0.3829
       PassThrough = -0.5045
        ---
       1best avg score: 16.34 (+16.34)
 1best avg model score: +12.6
         avg # updates: 1823
   non-0 feature count: 2883
           avg f count: 160
           avg list sz: 100
(time 0.2 min, 4 s/S)

---
Best iteration: 1 [GOLD = 16.34].
This took 0.2 min.