Science Question Answering on SQA-I (Accuracy)
67.9AccuracyOFA
Evaluation results
| Method | Links | |
|---|---|---|
| OFABackbone=LLaVA-v1.5-7B, Fine-tuning Dataset=Vision-Flan-186K, Sampling Ratio=15%2026.05 | 67.9 | |
| PreSelBackbone=LLaVA-v1.5-7B, Fine-tuning Dataset=Vision-Flan-186K, Sampling Ratio=15%2026.05 | 66.2 | |
| CoIDOBackbone=LLaVA-v1.5-7B, Fine-tuning Dataset=Vision-Flan-186K, Sampling Ratio=15%2026.05 | 66.2 | |
| COINCIDEBackbone=LLaVA-v1.5-7B, Fine-tuning Dataset=Vision-Flan-186K, Sampling Ratio=15%2026.05 | 63.9 | |
| EL2NBackbone=LLaVA-v1.5-7B, Fine-tuning Dataset=Vision-Flan-186K, Sampling Ratio=15%2026.05 | 62.7 | |
| CLIP-ScoreBackbone=LLaVA-v1.5-7B, Fine-tuning Dataset=Vision-Flan-186K, Sampling Ratio=15%2026.05 | 62.3 | |
| RandomBackbone=LLaVA-v1.5-7B, Fine-tuning Dataset=Vision-Flan-186K, Sampling Ratio=15%2026.05 | 62.1 | |
| ICONSBackbone=LLaVA-v1.5-7B, Fine-tuning Dataset=Vision-Flan-186K, Sampling Ratio=15%2026.05 | 60.2 | |
| Self-FilterBackbone=LLaVA-v1.5-7B, Fine-tuning Dataset=Vision-Flan-186K, Sampling Ratio=15%2026.05 | 59.3 | |
| TypiClustBackbone=LLaVA-v1.5-7B, Fine-tuning Dataset=Vision-Flan-186K, Sampling Ratio=15%2026.05 | 59.2 | |
| XMASBackbone=LLaVA-v1.5-7B, Fine-tuning Dataset=Vision-Flan-186K, Sampling Ratio=15%2026.05 | 58.1 | |
| IFDBackbone=LLaVA-v1.5-7B, Fine-tuning Dataset=Vision-Flan-186K, Sampling Ratio=15%2026.05 | 57.8 | |
| Full-DataBackbone=LLaVA-v1.5-7B, Fine-tuning Dataset=Vision-Flan-186K, Sampling Ratio=100%2026.05 | 55.6 |