moses-smt/mosesdecoder
1
1
Fork 0
mirror of https://github.com/moses-smt/mosesdecoder.git synced 2024-11-13 10:40:52 +03:00
Code Issues Packages Projects Releases Wiki Activity

Correction to calc. of lambda-gradient when more than 2 ttables

Browse Source
This commit is contained in:
Barry Haddow 2013-04-03 16:13:01 +01:00
parent 74e313e67c
commit c5965b8587
1 changed files with 4 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
contrib/promix/train.py
View File

@ -269,6 +269,7 @@ class MixtureModelTrainer:
# TODO: Better vectorisation
grad_list = []
for i, sample in enumerate(interpolated):
print>>sys.stderr,interp, sample[0],sample[1]
f_A = np.sum(np.log(sample[0]), axis=0)
f_B = np.sum(np.log(sample[1]), axis=0)
grad_list.append(f_A - f_B)
@ -312,7 +313,9 @@ class MixtureModelTrainer:
#print "hyp0",np.sum(phrase * (df_by_dlambda(sample[0])), axis=0)
#print "q1", df_by_dlambda(sample[1])
#print "hyp1",np.sum(phrase * (df_by_dlambda(sample[1])), axis=0),"\n"
grad_list[k] = np.sum(phrase * (df_by_dlambda(sample[0]) - df_by_dlambda(sample[1])), axis = 0).flatten()
#TODO: Check if the sum is required here. With 4 ttables and 4 features
# it gives lhs as (12) and rhs as (4)
grad_list[k] = (phrase * (df_by_dlambda(sample[0]) - df_by_dlambda(sample[1]))).flatten()
#grad_list = np.vstack(grad_list)
return grad_list