Speech Question Answering on English High School Listening Test (EHSL)
50Long Speech AccuracyUni-MoE w/ MoE-Task4 (4E) + Aux loss
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Uni-MoE w/ MoE-Task4 (4E) + Aux lossModality=I,T,S,V, Training Task Strategy=MoE-Task4, Number of Experts=4, Auxiliary Balancing Loss=true2024.05 | 50 | 54 | |
| Uni-MoE w/ MoE-Task4 (8E)Modality=I,T,S,V, Training Task Strategy=MoE-Task4, Number of Experts=82024.05 | 48.67 | 46 | |
| Uni-MoE w/ MoE-Task4 (4E)Modality=I,T,S,V, Training Task Strategy=MoE-Task4, Number of Experts=42024.05 | 48 | 58.67 | |
| Uni-MoE w/ MoE-Task2 (4E)Modality=I,T,S,V, Training Task Strategy=MoE-Task2, Number of Experts=42024.05 | 42 | 48 | |
| Uni-MoE w/ MoE-Task4 (8E) + Aux lossModality=I,T,S,V, Training Task Strategy=MoE-Task4, Number of Experts=8, Auxiliary Balancing Loss=true2024.05 | 42 | 50 | |
| Uni-MoE w/ MoE-Task1 (4E)Modality=I,T,S, Training Task Strategy=MoE-Task1, Number of Experts=42024.05 | 41.33 | 36 | |
| Single-Modality-Expert-Task6Modality=I,T,S, Training Task Strategy=MoE-Task62024.05 | 18.67 | 8 | |
| Single-Modality-Expert-Task5Modality=I,T,S, Training Task Strategy=MoE-Task52024.05 | 9.33 | 12 | |
| Macaw-LLMModality=Image, Text, Audio, Video2024.05 | 0.67 | 2 | |
| X-InstructBLIPModality=Image, Text, Audio, Video2024.05 | 0.67 | 2 |