Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Spell Correction on Twitter dataset (test)
Loading...
67.6
Whole-Word Accuracy
Seq2Seq (Soft-T2T)
45.448
51.199
56.95
62.701
Oct 7, 2019
Whole-Word Accuracy
Updated 4d ago
Evaluation Results
Method
Method
Links
Whole-Word Accuracy
Seq2Seq (Soft-T2T)
Architecture=Transform...
2019.10
67.6
PIE
Architecture=4 layer s...
2019.10
67
Ribeiro2018 (best model)
Model Type=local trans...
2019.10
54.1
Seq2Seq (Hard, lstm)
Attention=hard monoton...
2019.10
52.2
Seq2Seq (Soft, lstm)
Attention=soft, Archit...
2019.10
46.3
Feedback
Search any
task
Search any
task