Skip to content

Issues: ldmt-muri/alignment-with-openfst

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Clear current search query, filters, and sorts
62 Open 49 Closed
62 Open 49 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

control how much of the empirical probability mass of p(a,t|s) goes to U (a|t,s) urgent
#29 opened Nov 29, 2012 by wammar naacl deadline
@wammar
read the corpus to memory and permute it at the beginning of each iteration enhancement
#9 opened Nov 16, 2012 by wammar naacl deadline
@wammar
refactoring: in LogLinearModel::Train() the local variable gradient (as well as appliedL1Penalties) does not need to be an instance of LogLinearParams. It doesn't make much of a difference memory-wise, but it will be less confusing. enhancement
#12 opened Nov 16, 2012 by wammar
@wammar
refactor LogLinearModel::Train()
#13 opened Nov 16, 2012 by wammar
@wammar
shrink the discriminative lexicon as in chris et al. 2011 enhancement
#16 opened Nov 17, 2012 by wammar naacl deadline
@wammar
fill a grid of AER/perplexity results for czech urgent
#19 opened Nov 17, 2012 by wammar naacl deadline
@wammar
a particle filter instead of the importance sampler urgent
#20 opened Nov 17, 2012 by wammar naacl deadline
@wammar
introduce representation variables
#22 opened Nov 19, 2012 by wammar
@wammar
tokenize english and czech training data to look like the AER testset bug
#24 opened Nov 19, 2012 by wammar naacl deadline
@wammar
add command line arguments to train-loglinear which specify (hyper)parameters urgent
#25 opened Nov 19, 2012 by wammar naacl deadline
@wammar
compare the two implementations of chris et al 2011
#26 opened Nov 20, 2012 by wammar naacl deadline
@wammar
HMM likelihood >>> model 1 likelihood. make sure both are properly normalized bug
#27 opened Nov 29, 2012 by wammar naacl deadline
@wammar
use multiple "debug levels" to decide when to print debugging info enhancement
#3 opened Oct 24, 2012 by wammar
@wammar
model 1 gives -ve loglikelihood for the tiny dataset in both directions bug
#28 opened Nov 29, 2012 by wammar naacl deadline
@wammar
tsuruoka's l1 trick might be buggy. bug
#30 opened Nov 29, 2012 by wammar naacl deadline
@wammar
precompute ortho-similiarity features enhancement
#31 opened Nov 29, 2012 by wammar naacl deadline
@wammar
replace map<string,*> with a trie #urgent urgent
#35 opened Nov 29, 2012 by wammar naacl deadline
@wammar
use alias sampling in hmm aligner
#37 opened Dec 3, 2012 by wammar naacl deadline
@wammar
add an L2 regularizer for the multinomials latent-crf
#40 opened Dec 13, 2012 by wammar
@wammar
update the write-up latent-crf
#46 opened Dec 26, 2012 by wammar
@wammar
read matt's draft latent-crf
#47 opened Dec 26, 2012 by wammar
@wammar
read Desai's "Unsupervised Bilingual POS Tagging with Markov Random Fields" http://www.cs.cmu.edu/~cdyer/acl10mrfpos.pdf latent-crf
#51 opened Dec 27, 2012 by wammar
@wammar
enforce certain constraints by freezing features latent-crf
#54 opened Dec 28, 2012 by wammar
@wammar
2
read Avi's Exact and Efficient Parallel Inference for NonParametric Mixture Models
#56 opened Jan 3, 2013 by wammar
@wammar
synchronization of access to objects in shared memory is error prone
#111 opened Nov 14, 2013 by wammar
ProTip! Updated in the last three days: updated:>2025-01-05.