issues
search
tuetschek
/
ratpred
Referenceless quality estimation for natural language generation – predicting human ratings
Other
8
stars
0
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Improper masking on ranking instances
#13
tuetschek
closed
5 years ago
2
Updating to current version of TGen
#12
tuetschek
closed
7 years ago
0
Try enc-dec with forced decoding
#11
tuetschek
closed
7 years ago
1
Pretrain by language modeling
#10
tuetschek
opened
7 years ago
2
More sophisticated error generation
#9
tuetschek
opened
7 years ago
0
Pretrain by learning to generate?
#8
tuetschek
closed
7 years ago
2
Fabricate data from system outputs
#7
tuetschek
closed
7 years ago
1
Tougher delexicalization
#6
tuetschek
closed
7 years ago
1
Cross-validation -- compute actual scores based on ratings
#5
tuetschek
closed
7 years ago
2
measure accuracy as well as distance
#4
tuetschek
closed
7 years ago
1
finish score grepping & averaging scripts
#3
tuetschek
closed
7 years ago
1
add output prediction option for testing
#2
tuetschek
closed
7 years ago
1
add random seed option for training
#1
tuetschek
closed
7 years ago
1