diff options
| author | Patrick Simianer <p@simianer.de> | 2015-01-23 15:50:27 +0100 |
|---|---|---|
| committer | Patrick Simianer <p@simianer.de> | 2015-01-23 15:50:27 +0100 |
| commit | 32dea3f24e56ac7c17343457c48f750f16838742 (patch) | |
| tree | 79177b58cbff08c14991a0da8e851912b1c06309 /training/dtrain/examples/toy/expected-output | |
| parent | 556dc935c7a2d8df78a35447d20d71b4bf6e391a (diff) | |
dtrain: multi-reference BLEU
Diffstat (limited to 'training/dtrain/examples/toy/expected-output')
| -rw-r--r-- | training/dtrain/examples/toy/expected-output | 31 |
1 files changed, 24 insertions, 7 deletions
diff --git a/training/dtrain/examples/toy/expected-output b/training/dtrain/examples/toy/expected-output index 1da2aadd..fbee24e3 100644 --- a/training/dtrain/examples/toy/expected-output +++ b/training/dtrain/examples/toy/expected-output @@ -1,26 +1,29 @@ Warning: hi_lo only works with pair_sampling XYX. cdec cfg 'cdec.ini' -Seeding random number sequence to 1664825829 +Seeding random number sequence to 3626026233 dtrain Parameters: k 4 N 4 T 2 + batch 0 scorer 'bleu' sample from 'kbest' filter 'uniq' learning rate 1 gamma 0 loss margin 0 + faster perceptron 1 pairs 'all' pair threshold 0 select weights 'last' l1 reg 0 'none' + pclr no max pairs 4294967295 + repeat 1 cdec cfg 'cdec.ini' - input 'src' - refs 'tgt' + input '' output '-' (a dot represents 10 inputs) Iteration #1 of 2. @@ -32,12 +35,19 @@ WEIGHTS small_rule = -2 little_rule = +3 PassThrough = -5 + PassThrough_1 = +0 + PassThrough_2 = +0 + PassThrough_3 = +0 + PassThrough_4 = +0 + PassThrough_5 = +0 + PassThrough_6 = +0 --- 1best avg score: 0.5 (+0.5) 1best avg model score: 2.5 (+2.5) - avg # pairs: 4 - avg # rank err: 1.5 + avg # pairs: 1.5 + avg # rank err: 1.5 (meaningless) avg # margin viol: 0 + k-best loss imp: 100% non0 feature count: 6 avg list sz: 4 avg f count: 2.875 @@ -52,12 +62,19 @@ WEIGHTS small_rule = -2 little_rule = +3 PassThrough = -5 + PassThrough_1 = +0 + PassThrough_2 = +0 + PassThrough_3 = +0 + PassThrough_4 = +0 + PassThrough_5 = +0 + PassThrough_6 = +0 --- 1best avg score: 1 (+0.5) 1best avg model score: 5 (+2.5) - avg # pairs: 5 - avg # rank err: 0 + avg # pairs: 0 + avg # rank err: 0 (meaningless) avg # margin viol: 0 + k-best loss imp: 100% non0 feature count: 6 avg list sz: 4 avg f count: 3 |