Image Captioning on Falcon-X (test)
40.92BLEUFalcon
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| FalconEvaluation Protocol=Fine-tuned2026.06 | 40.92 | 47.74 | 0.064 | |
| GromaEvaluation Protocol=Fine-tuned2026.06 | 40.7 | 47.69 | 0.061 | |
| LISAEvaluation Protocol=Fine-tuned2026.06 | 35.57 | 39.13 | 0.04 | |
| Sting-BeeEvaluation Protocol=Fine-tuned2026.06 | 33.38 | 37.48 | 0.024 | |
| Sa2VaEvaluation Protocol=Zero-shot2026.06 | 25.06 | 18.11 | 0.012 | |
| Sting-BeeEvaluation Protocol=Zero-shot2026.06 | 24.61 | 17.95 | 0.012 | |
| LISAEvaluation Protocol=Zero-shot2026.06 | 20.57 | 18.08 | 0.01 | |
| GromaEvaluation Protocol=Zero-shot2026.06 | 18.8 | 16.48 | 0.004 | |
| Kosmos-2Evaluation Protocol=Zero-shot2026.06 | 16.26 | 17.56 | 0.002 | |
| GLaMMEvaluation Protocol=Zero-shot2026.06 | 15.91 | 17.56 | 0.0032 | |
| Llava-1.5Evaluation Protocol=Zero-shot2026.06 | 12.67 | 15.67 | — |