h100

SSH into h100_sglang to attach to the sglang_bbuf container for GPU-enabled SGLang development.

721|65|Updated Apr 1, 2026

Search Tags:#gpu #container #ssh #cuda #diffusion #sglang #remote-development

Ready to run

npx skills add https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS --skill h100

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: h100
Source: https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS/tree/main/skills/h100
Command: npx skills add https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS --skill h100

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

SSH into the H100 remote environment to enable GPU-backed SGLang development when local hardware lacks CUDA resources, using the preconfigured container and workspace for reproducible runs.

Core Features & Use Cases

Remote GPU-enabled development: access h100_sglang, attach to the sglang_bbuf container, and work in /sgl-workspace/sglang.
CUDA and diffusion validation: run GPU-bound tests, smoke checks, and remote validation safely on the H100 box.
Reproducible workflows: use a ready remote environment to ensure consistent results across teams.

Quick Start

Connect to h100_sglang, attach to the sglang_bbuf container, and begin GPU-enabled development in /sgl-workspace/sglang.

Frequently Asked Questions about h100

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I run SGLang development on a remote GPU when my local machine lacks CUDA resources?▼

Remote GPU-enabled SGLang development solves local CUDA hardware limitations by SSHing into an H100 host, attaching to a preconfigured Docker container, and accessing the mounted SGLang repository workspace for reproducible runs.

Can I run diffusion checks and CUDA smoke tests on remote H100 hardware?▼

Yes, CUDA and diffusion validation can be executed on remote H100 hardware by connecting to the host via SSH, attaching to the sglang_bbuf Docker container, and running GPU-bound smoke tests safely in the isolated environment.

What do I need to connect to an H100 container for remote GPU development?▼

Remote GPU development requires SSH access to the h100_sglang host, a running Docker container named sglang_bbuf, and the SGLang repository mounted at /sgl-workspace/sglang to ensure a ready and reproducible environment.

What is the best way to ensure reproducible CUDA workflows across teams without local GPUs?▼

Using a ready remote H100 environment ensures reproducible CUDA workflows across teams by providing a preconfigured Docker container and mounted SGLang repository, guaranteeing consistent results for GPU-bound tests and validation.

Why should I use a remote Docker container for SGLang validation instead of local hardware?▼

A remote Docker container provides H100 GPU access for SGLang validation when local hardware is insufficient, ensuring safe execution of diffusion checks and CUDA tests in a standardized, reproducible workspace.

Does remote SGLang development on an H100 support working directly in the repository workspace?▼

Yes, remote SGLang development supports working directly in the repository by attaching to the sglang_bbuf container where the SGLang repository is mounted at /sgl-workspace/sglang, enabling immediate GPU-backed development.