skills.rest
Clay-HHK
Clay-HHKCOMMUNITY
Explore all skills by this author →
GitHub

lambda-labs-gpu-cloud

Launch and manage Lambda Labs GPU cloud instances for ML training and inference.

1|Updated Mar 5, 2026
Search Tags:#infrastructure#ssh#distributed training#slurm#lambda labs#gpu cloud#llm fine-tuning
One-click install
npx skills add https://github.com/Clay-HHK/claude-skills --skill lambda-labs-gpu-cloud-clay-hhk
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: lambda-labs-gpu-cloud
Source: https://github.com/Clay-HHK/claude-skills/tree/main/lambda-labs
Command: npx skills add https://github.com/Clay-HHK/claude-skills --skill lambda-labs-gpu-cloud-clay-hhk

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires lambda-cloud-client, and includes references (resource) components.

What problem does it solve? It removes the operational overhead of provisioning and managing dedicated GPU cloud infrastructure, letting you launch, configure, and run ML training or inference workloads on Lambda Labs instances with SSH access, persistent storage, and multi-node clusters. ## Core Features & Use Cases - Instance Lifecycle Management: Launch, list, and terminate GPU instances (B200, H100, A100, A10, and more) via the Python API, curl CLI, or web console. - Distributed Training Workflows: Run single-node multi-GPU jobs with torchrun/DDP and multi-node jobs on 1-Click Slurm clusters with InfiniBand, FSDP, or DeepSpeed. - Persistent Storage & Cost Control: Attach filesystems for checkpoints and datasets, choose cost-appropriate GPUs per workload, and avoid data loss on termination. - Use Case: Fine-tune a 70B LLM by launching an 8x H100 instance with an attached filesystem, running accelerate across all GPUs, and saving checkpoints that survive instance termination. ## Quick Start Ask the assistant to launch a Lambda Labs H100 instance with your SSH key and filesystem attached, then walk you through connecting and starting a training job.

Frequently Asked Questions about lambda-labs-gpu-cloud

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

FAQPage Schema
How do I launch a Lambda Labs GPU instance with the Python API?▼

Install lambda-cloud-client, configure it with your API key, and call launch_instance with a region, instance type name like gpu_1x_h100_sxm5, and your SSH key names. The instance typically becomes active in 3-15 minutes depending on GPU count.

Which Lambda Labs GPU should I choose for LLM fine-tuning?▼

For 7B models, a single A100 40GB at $1.29/hr works well. For 70B models, use an 8x H100 SXM instance. For inference or development, A10 or A6000 instances offer lower hourly costs.

Lambda Labs vs Modal vs RunPod for GPU workloads?▼

Lambda Labs suits dedicated instances with full SSH access, long training jobs, and multi-node InfiniBand clusters. Modal fits serverless auto-scaling workloads, while RunPod offers cheaper spot instances and serverless endpoints.

Why did I lose my data after terminating a Lambda instance?▼

The root volume at /home/ubuntu is ephemeral and deleted on termination. Attach a persistent filesystem at launch time and store datasets, checkpoints, and outputs under /lambda/nfs/<filesystem_name> to keep them.

How do I fix SSH permission denied on a Lambda instance?▼

Verify the SSH key was added to the Lambda console before launching, since keys cannot be attached afterward. Check key file permissions with chmod 600 on the private key, and confirm the instance status is active.

How do I run multi-node distributed training on Lambda 1-Click Clusters?▼

Submit a Slurm job with srun or sbatch specifying nodes and GPUs per node, then launch torchrun with the c10d rendezvous backend using the master node address. The clusters include InfiniBand, NCCL, and pre-installed PyTorch with DDP and FSDP.

Related Skills

Explore More Software Engineering Skills→
KarlinskySKarlinskyS

lambda-labs-gpu-cloud

Provision Lambda Labs GPU cloud instances for machine learning training and inference.

Community
Intermediate
RachasumanthRachasumanth

inference-deploy

Deploy trained machine learning models for inference using vLLM, TGI, Ollama, and llama.cpp.

Community
Advanced
dogekikidogekiki

monitor-experiment

Monitor remote machine learning experiment status and performance metrics across SSH, Vast.ai, and Modal.

Community
Advanced
hve4638hve4638

serverless-modal

Automate GPU workload deployment and orchestration on Modal's serverless platform.

Community
Advanced