Autor: Jakub Rusinowski · Ostatnia aktualizacja: 8 października 2024
Najlepszy otwarty wielojęzykowy model na świecie. Aya Expanse obsługuje 23 języki i jest specjalnie zoptymalizowany dla użytkowników nie-anglojęzycznych. Przewyższa Llama 3.1 w arabskim, chińskim, hindi, francuskim, hiszpańskim, rosyjskim i 17 innych językach.
| Licence | What it permits | Applies to |
|---|---|---|
CC-BY-NC-4.0 | Research / non-commercial only Research / non-commercial only — this licence does NOT permit shipping a commercial product. | Aya Expanse 8B, Aya Expanse 32B |
| Aya Expanse 8B | Min 6 GB VRAM · Q4_K_M · 8,192 ctx · ollama run aya-expanse:8b |
| Aya Expanse 32B | Min 20 GB VRAM · Q4_K_M · 128,000 ctx · ollama run aya-expanse:32b |
The cheapest GPU that runs Aya Expanse locally (min 6 GB VRAM) is the Intel Arc B570 (10 GB).
Install Ollama then run: ollama run aya-expanse:8b
Minimum VRAM: 6 GB. For best results use Q4_K_M quantization.
Aya Expanse needs about 6 GB VRAM at Q4_K_M quantization for its smallest variant. Variants: Aya Expanse 8B (6 GB, Q4_K_M); Aya Expanse 32B (20 GB, Q4_K_M). On Apple Silicon, unified memory counts toward this requirement.
Yes — Aya Expanse 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 Aya Expanse 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 aya-expanse:8b. This downloads Aya Expanse and starts a local, OpenAI-compatible endpoint — no internet connection is needed after the initial download.