Tranx

Reported on 5 benchmarks across 2 tasks · 1 paper · 2 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Natural Language Processing5 results

Code GenerationonDjango
Accuracy· 2018-10-05
73.7
best: 81.83 (MarianCG)
SOTA
TRANX: A Transition-based Neural Abstract Syntax Parser for Semantic Parsing and Code Generation arXiv:1810.02720
Semantic ParsingonATIS
Accuracy· 2018-10-05
86.2
best: 90.4 (PhraseTransformer)
SOTA
TRANX: A Transition-based Neural Abstract Syntax Parser for Semantic Parsing and Code Generation arXiv:1810.02720
Code GenerationonWikiSQL
Exact Match Accuracy· 2018-10-05
68.6
best: 83.7 (NL2SQL-RULE)
TRANX: A Transition-based Neural Abstract Syntax Parser for Semantic Parsing and Code Generation arXiv:1810.02720
Code GenerationonWikiSQL
Execution Accuracy· 2018-10-05
78.6
best: 89.2 (NL2SQL-RULE)
TRANX: A Transition-based Neural Abstract Syntax Parser for Semantic Parsing and Code Generation arXiv:1810.02720
Semantic ParsingonGeo
Accuracy· 2018-10-05
87.7
best: 88.2 (coarse2fine)
TRANX: A Transition-based Neural Abstract Syntax Parser for Semantic Parsing and Code Generation arXiv:1810.02720