Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

DSTC2

Benchmarks

Task NameDataset NameSOTA ResultTrend
Dialogue State TrackingDSTC2 (test)
Joint Goal Accuracy85
39
Dialog GenerationDSTC2 (test)
Accuracy (Response)47.4
10
Dialogue act predictionDSTC2
Micro-F1 Score94.6
7
DialogDSTC2
Average Error Rate0.489
7
Dialogue act predictionDSTC2 (10% Data)
Micro F1 Score93.6
6
Dialogue act predictionDSTC2 (1% Data)
Micro F183.7
6
Dialogue act predictionDSTC2 (Full Data)
Micro-F1-
0
Response SelectionDSTC2
1-to-100 Accuracy-
0
Showing 8 of 8 rows