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
****************************************************************************************************
dtrain
Parameters:
                       k 100
                       N 4
                       T 1
           learning rate 0.0001
            error margin 1
                  l1 reg 0
            decoder conf 'cdec.ini'
                   input 'work/shard.3.0.in'
                  output 'work/weights.3.2'
              weights in 'work/weights.1'
(a dot per input)
Iteration #1 of 1.
 .. 1
WEIGHTS
              Glue = -0.12993
       WordPenalty = +0.13651
     LanguageModel = +0.58946
 LanguageModel_OOV = -0.48362
     PhraseModel_0 = -0.81262
     PhraseModel_1 = +0.44273
     PhraseModel_2 = +1.1733
     PhraseModel_3 = -0.1826
     PhraseModel_4 = -1.2213
     PhraseModel_5 = +0.02435
     PhraseModel_6 = -0.18823
       PassThrough = -0.51378
        ---
       1best avg score: 0.12674 (+0.12674)
 1best avg model score: -7.2878
           avg # pairs: 1769
   non-0 feature count: 12
           avg list sz: 100
           avg f count: 12
(time 0.1 min, 6 s/S)

---
Best iteration: 1 [GOLD = 0.12674].
This took 0.1 min.