Closed-ended Task T-5 on DanmakuTPP-QA
0.1RMSEMM-TPP
Evaluation results
| Method | Links | |
|---|---|---|
| MM-TPPModel Category=Finetuned2026.02 | 0.1 | |
| Qwen2.5-VL-3BModel Category=Finetuned2026.02 | 0.16 | |
| DeepSeek-V3Model Category=LLMs2026.02 | 0.26 | |
| Llama-3.3-70B-InstructModel Category=LLMs2026.02 | 0.27 | |
| Gemma3-27BModel Category=MLLMs2026.02 | 0.27 | |
| Qwen2.5-32B-InstructModel Category=LLMs2026.02 | 0.29 | |
| Qwen3-235B-A22BModel Category=LLMs2026.02 | 0.31 | |
| Qwen3-30B-A3BModel Category=LLMs2026.02 | 0.33 | |
| Qwen3-8BModel Category=LLMs2026.02 | 0.41 | |
| Qwen2.5-72B-InstructModel Category=LLMs2026.02 | 0.46 | |
| Qwen2.5-VL-72BModel Category=MLLMs2026.02 | 0.47 | |
| Qwen2.5-VL-32BModel Category=MLLMs2026.02 | 0.51 | |
| Qwen2.5-7B-InstructModel Category=LLMs2026.02 | 0.56 | |
| Qwen2.5-VL-7BModel Category=MLLMs2026.02 | 0.82 |