MultiWOZ
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
MultiWOZ 2.1 (test)
99.62Inform Rate
11
MultiWOZ 2.3 (test)
63JGA
11
MultiWOZ 2.1 (5%)
44.98Joint Goal Acc
11
MultiWOZ 2.1 (1%)
37.26Joint Goal Acc
10
MultiWOZ 2.2 (test)
90Inform
10
MultiWOZ 20% 2.0 (train)
90.25Inform
10
MultiWOZ 2.0 (10% train)
86.3Inform Rate
10
MultiWOZ 5% 2.0 (train)
85.65Inform
10
MultiWOZ Taxi domain 1.0
106.3Combined Score
10
MultiWOZ 1.0 (train)
104.9Combined Score
10
MultiWOZ Attraction domain 1.0
96Combined Score
10
MultiWOZ Hotel domain 1.0
100.7Combined Score
10
MultiWOZ Restaurant domain 1.0
97.6Combined Score
10
MultiWOZ
56Accuracy
9
MultiWOZ low-resource 2.1
29.52Average Precision (AP)
9
MultiWOZ 2.1
51.24Success Rate (SR)
9
MultiWOZ 2.2
88.3Success Rate
9
MultiWOZ 2.3
63.1Joint Goal Accuracy
9
MultiWOZ 2.2
62.3Joint GA
9
MultiWOZ 2.3
35.7Exact Match (EM)
8
MultiWOZ
20.6Backward Transfer (BWT)
8
MultiWOZ 2.4
77.3Joint Goal Accuracy
8
MultiWOZ 2.2 (test)
0.379JGA
8
MultiWoZ (test)
79.9Rouge-L
8
MultiWOZ 1.0 (test)
13.7RAG Calls
7