List your company

Resource Utilization on MSR-VTT (Video-Language Understanding)

3,751Initial Memory (MB)Video-LLaMA3
2,776.889,352.1915,927.522,502.81Apr 9, 2026

Evaluation results

MethodLinks
Video-LLaMA3Size=2B2026.043,75126,70522,95533.5
InternVL2.5Size=2.2B2026.044,59312,6988,10531.5
ABMambaSize=3.6B2026.047,0887,57048295.4
Video-ChatGPTSize=7B2026.0413,44015,5052,06638.1
LLaVA-OneVisionSize=7B2026.0415,81321,0045,19124.8
Video-LLaVASize=7B2026.0428,10430,6522,54828.9