Autor: Jakub Rusinowski · Ostatnia aktualizacja: 25 września 2024
Multimodalne rozszerzenie Meta Llama 3.2 z enkoderem wizji do rozumienia obrazów. Wersja 11B działa na jednym GPU 8–12 GB VRAM. Obsługuje analizę obrazów, OCR, odczytywanie wykresów i wizualne pytania-odpowiedzi.
| Licence | What it permits | Applies to |
|---|---|---|
Llama 3.2 Community | Commercial use permitted Weights are downloadable and commercial use is permitted, subject to the licence’s acceptable-use terms. | Llama 3.2 Vision 11B, Llama 3.2 Vision 90B |
| Llama 3.2 Vision 11B | Min 7 GB VRAM · Q4_K_M · 128,000 ctx · ollama run llama3.2-vision:11b |
| Llama 3.2 Vision 90B | Min 54 GB VRAM · Q4_K_M · 128,000 ctx · ollama run llama3.2-vision:90b |
The cheapest GPU that runs Llama 3.2 Vision locally (min 7 GB VRAM) is the Intel Arc B570 (10 GB).
Install Ollama then run: ollama run llama3.2-vision:11b
Minimum VRAM: 7 GB. For best results use Q4_K_M quantization.
Llama 3.2 Vision needs about 7 GB VRAM at Q4_K_M quantization for its smallest variant. Variants: Llama 3.2 Vision 11B (7 GB, Q4_K_M); Llama 3.2 Vision 90B (54 GB, Q4_K_M). On Apple Silicon, unified memory counts toward this requirement.
Yes — Llama 3.2 Vision runs on an RTX 4090 (24 GB) and other 24 GB cards such as the RTX 3090. Smaller variants also fit comfortably on 8–16 GB GPUs at Q4_K_M.
Q4_K_M is the best balance of quality and VRAM for Llama 3.2 Vision in most cases. Choose Q8_0 for near-lossless quality if you have spare VRAM, or smaller quants (Q3/Q2) only when memory is tight.
Install Ollama, then run: ollama run llama3.2-vision:11b. This downloads Llama 3.2 Vision and starts a local, OpenAI-compatible endpoint — no internet connection is needed after the initial download.