Chat
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
chat 1m (test)
100MCA Accuracy
21
Chat
49.3Chat Score
8
CHAT
98.4AUC
7
Chat 1m-Conv
99MCA
6
Chat 1m
100MCA
6
Chat (test)
28.23ROUGE-1
6
chat 1m
0.01Inference Latency (per prompt)
5
chat in-distribution (test)
86Macro F1 (Mean)
4
Chat
57Win Rate
3
CHAT
86Macro F1
2