Video Captioning on VDC aggregate
0.539LLM Judge ScoreSmolVLM
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| SmolVLMModel Scale=2.2B, Evaluation Protocol=LLM Judge2026.06 | 0.539 | — | — | — | — | — | |
| SmolVLMModel Scale=500M, Evaluation Protocol=LLM Judge2026.06 | 0.515 | — | — | — | — | — | |
| VIDEO2LORAModel Scale=2.2B, Evaluation Protocol=LLM Judge2026.06 | 0.511 | — | -0.028 | -0.037 | — | — | |
| VIDEO2LORAModel Scale=500M, Evaluation Protocol=LLM Judge2026.06 | 0.406 | — | -0.108 | -0.118 | — | — | |
| SmolVLMModel Scale=500M, Evaluation Protocol=Token F12026.06 | — | 31.5 | — | — | — | — | |
| SmolVLMModel Scale=2.2B, Evaluation Protocol=Token F12026.06 | — | 29.7 | — | — | — | — | |
| VIDEO2LORAModel Scale=500M, Evaluation Protocol=Token F12026.06 | — | 28.8 | -0.027 | -0.03 | — | — | |
| VIDEO2LORAModel Scale=2.2B, Evaluation Protocol=Token F12026.06 | — | 30.4 | 0.007 | 0.003 | — | — |