Cloud Gpu LLM

Run local LLMs (llama.cpp, GGUF) on a rented spot GPU VM on Google Cloud when a laptop is too slow or when you need a benchmark. Covers the gpu-vm script, the IAP SSH setup, the traps that cost time in a real run, cost, and safe access through a tunnel. Use when a local model reads prompts too slowly, when asked to benchmark a model on an A100, L4 or H100, or when asked to run llama-server in the cloud.

pooriaarab Updated 1 repo stars

File contents

pooriaarab/skills commit 4e9ad0f191

Frequently asked questions

npx get-skill add pooriaarab/cloud-gpu-llm