51 lines
1.2 KiB
Markdown
51 lines
1.2 KiB
Markdown
---
|
|
layout: default
|
|
---
|
|
|
|
<h1 class="!text-2xl">Ollama: Requisitos de Hardware</h1>
|
|
|
|
<div class="text-sm mt-2">
|
|
|
|
<h2 class="!text-base">📊 ¿Qué necesitas según el modelo?</h2>
|
|
|
|
<div class="mt-2 text-xs">
|
|
|
|
| Tamaño | RAM | VRAM GPU | Modelos Populares |
|
|
|--------|-----|----------|-------------------|
|
|
| **3B** | 8GB | 4GB | Llama 3.2 3B, Phi-3 Mini |
|
|
| **7B** | 16GB | 8GB | Llama 3.2 7B, Mistral 7B |
|
|
| **13B** | 32GB | 16GB | Llama 3.2 13B |
|
|
| **70B** | 64GB+ | 24GB+ | Llama 3.3 70B |
|
|
|
|
</div>
|
|
|
|
<v-click>
|
|
|
|
<h2 class="!text-base">💡 Recomendaciones</h2>
|
|
|
|
<div class="grid grid-cols-3 gap-3 mt-2">
|
|
|
|
<div class="border-l-4 border-green-500 pl-3">
|
|
<p class="font-bold text-sm">🎓 Para empezar</p>
|
|
<p class="text-xs mt-1">Llama 3.2 3B</p>
|
|
<p class="text-xs opacity-70">8GB VRAM suficiente</p>
|
|
</div>
|
|
|
|
<div class="border-l-4 border-blue-500 pl-3">
|
|
<p class="font-bold text-sm">💻 Desarrollo</p>
|
|
<p class="text-xs mt-1">Llama 3.2 7B</p>
|
|
<p class="text-xs opacity-70">16GB VRAM + GPU</p>
|
|
</div>
|
|
|
|
<div class="border-l-4 border-purple-500 pl-3">
|
|
<p class="font-bold text-sm">🏢 Producción</p>
|
|
<p class="text-xs mt-1">Usa APIs cloud</p>
|
|
<p class="text-xs opacity-70">Más fiable</p>
|
|
</div>
|
|
|
|
</div>
|
|
|
|
</v-click>
|
|
|
|
</div>
|