# Video LLM FT: LLaVA-NeXT-Video + VideoLLaMA3 + Frame Sampling Stratejisi

> Source: https://sukruyusufkaya.com/learn/fine-tuning-cookbook/ftc-video-llm-finetuning
> Updated: 2026-08-20T15:46:26.344Z
> Category: Fine-Tuning Cookbook (Model-by-Model)
> Module: Part VI — Vision-Language Multimodal FT
**TLDR:** Video LLM'i — image'in temporal extension'ı. LLaVA-NeXT-Video, VideoLLaMA3, Qwen 2.5-VL native video. Frame sampling (uniform vs adaptive), temporal token compression, long-video Q&A (>1 saat). RTX 4090'da Video LLM FT — short-clip (10-30 sn) ile pratik.