Seongwoo Choi / Journal

Run vLLM in Docker on an NVIDIA GPU, step by step

You need
A Linux machine with an NVIDIA GPU and its driver, Docker, and sudo rights.
You end up with
vLLM running as a container that restarts with the machine, answering on a local port.

Serve Qwen with vLLM: chat, tool calling and thinking

You need
The Docker setup from part 2 and one NVIDIA GPU with 24 GB of memory or more.
You end up with
A Qwen3 server that chats, shows or hides its reasoning, and calls your functions.

Serve Phind-CodeLlama-34B on A100 GPUs with vLLM

You need
The Docker setup from part 2, about 70 GB of free disk, and one A100 80GB or two A100 40GB cards.
You end up with
A private coding assistant behind an OpenAI-style chat endpoint, with a template you wrote and tested.

vLLM for a team: batching, API keys and a safe proxy

You need
A working vLLM container from parts 2 to 4, a host name for the server, and a TLS certificate for it.
You end up with
One HTTPS endpoint where each person has their own key, a rate limit and a usage log, and nothing else is reachable.

Monitor vLLM with Prometheus, Grafana and DCGM in Docker

You need
The GPU host and Docker setup from part 2. About 2 GB of disk for images and data.
You end up with
Dashboards for CPU, memory, disk, every GPU and the model server, plus alert rules, all started by one command.

vLLM alerts with Alertmanager: know before your users do

You need
The monitoring stack from part 7, and a Telegram account. The test in section 6 needs only Prometheus, Alertmanager and Python on any machine.
You end up with
Urgent alerts that ring your phone, minor ones that arrive silently, a way to pause them during maintenance, and a check that the alerting itself is alive.

Mac mini as a personal AI assistant with Ollama

You need
A Mac mini with Apple silicon on macOS 14 or newer, 16 GB of memory or more, and your home network.
You end up with
A private assistant that is always on, answers in a browser on your phone and laptop, and speaks the OpenAI API to your scripts.

Claude Code hooks, tested: five scripts you can copy

You need
Claude Code, a project folder, bash and jq. On Windows, Git for Windows provides bash.
You end up with
Five working hooks in .claude/hooks, one settings file that registers them, and a script that proves each one behaves.

Claude Code plugins and MCP servers: a terminal guide

You need
Claude Code installed on macOS or Windows, a terminal, and Python 3 for the practice server.
You end up with
One remote and one local MCP server connected, one plugin installed, and the commands to manage both.

Build an MCP server in Python from scratch, step by step

You need
Python 3.10 or later and Claude Code. No GPU, no API key and no packages for the first version.
You end up with
A to-do list server that Claude Code connects to and can use through three tools, in two versions: hand-written and SDK-based.

Claude Code and your email: what to check this week

You need
Claude Code, bash and jq. On Windows, Git for Windows provides bash.
You end up with
A current version with the built-in protection on, one extra hook that blocks tool calls carrying your address, and a repository with no personal settings file in it.

Run Claude Code on your own vLLM server: what to expect

You need
A vLLM server with a model that supports tool calling, and Claude Code. Section 4's test needs only Python and Claude Code.
You end up with
Claude Code talking to a model on your own hardware, with a clear view of what works, what does not and what it costs in context.

Claude Code request size: what our local recorder measured

You need
Claude Code and Python. No API key and no model: the requests go to a recorder on your own machine.
You end up with
The size of your own setup, part by part, and a short list of what to trim.