Image Captioning 벤치마크
Image Captioning on Object HalBench
chair_i
- 2023-12-01 — RLHF-V: chair_i 7.5
| Rank | Model | chair_i | chair_s | Paper | Code | Year |
|---|---|---|---|---|---|---|
| 1 | RLHF-V | 7.5 | 12.2 | RLHF-V: Towards Trustworthy MLLMs via Behavior Alignment from Fine-grained Correctional Human Feedback | openbmb/minicpm-v · rlhf-v/rlhf-v · tidedra/vl-rlhf · +1 | 2023 |
| 2 | RLAIF-V 7B | 4.3 | 8.5 | RLAIF-V: Open-Source AI Feedback Leads to Super GPT-4V Trustworthiness | openbmb/minicpm-v · openbmb/omnilmm · OpenBMB/MiniCPM-o · +2 | 2024 |
| 3 | RLAIF-V 12B | 1.8 | 3.3 | RLAIF-V: Open-Source AI Feedback Leads to Super GPT-4V Trustworthiness | openbmb/minicpm-v · openbmb/omnilmm · OpenBMB/MiniCPM-o · +2 | 2024 |