Skip to main content
skypilot-multi-cloud-orchestration Multi-cloud orchestration for ML workloads with automatic cost optimization. Use when you need to run training or batch jobs across multiple clouds, leverage spot instances with auto-recovery, or optimize GPU costs across providers.
설치로 이동 Skills Marketplace 커뮤니티가 만든 AI 스킬을 발견하고 탐색하세요.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
직접 명령은 검토 Prompt를 거치지 않습니다. 실행하기 전에 소스를 확인하세요.
npx skills add https://github.com/davila7/claude-code-templates --skill skypilot-multi-cloud-orchestration명령은 한 줄로 유지됩니다. 복사하기 전에 가로로 스크롤해 전체 내용을 확인하세요.
로컬 사본을 원하시나요? SkillsMP에서 현재 제공할 수 있는 파일을 다운로드하세요.
Zip 다운로드 다운로드 중... 이 저장소의 다른 Skills Build AI agents that interact with computers like humans do - viewing screens, moving cursors, clicking buttons, and typing text. Covers Anthropic's Computer Use, OpenAI's Operator/CUA, and open-source alternatives. Critical focus on sandboxing, security, and handling the unique challenges of vision-based control. Use when: computer use, desktop automation agent, screen control AI, vision-based agent, GUI automation.
Expert in building products that wrap AI APIs (OpenAI, Anthropic, etc.) into focused tools people will pay for. Not just 'ChatGPT but different' - products that solve specific problems with AI. Covers prompt engineering for products, cost management, rate limiting, and building defensible AI businesses. Use when: AI wrapper, GPT product, AI tool, wrap AI, AI SaaS.
github-workflow-automation Automate GitHub workflows with AI assistance. Includes PR reviews, issue triage, CI/CD integration, and Git operations. Use when automating GitHub workflows, setting up PR review automation, creating GitHub Actions, or triaging issues.
name skypilot-multi-cloud-orchestration description Multi-cloud orchestration for ML workloads with automatic cost optimization. Use when you need to run training or batch jobs across multiple clouds, leverage spot instances with auto-recovery, or optimize GPU costs across providers. version 1.0.0 author Orchestra Research license MIT tags ["Infrastructure","Multi-Cloud","Orchestration","GPU","Cost Optimization","SkyPilot"] dependencies ["skypilot>=0.7.0"]
SkyPilot Multi-Cloud Orchestration
Comprehensive guide to running ML workloads across clouds with automatic cost optimization using SkyPilot.
When to use SkyPilot
Use SkyPilot when:
Running ML workloads across multiple clouds (AWS, GCP, Azure, etc.)
Need cost optimization with automatic cloud/region selection
Running long jobs on spot instances with auto-recovery
Managing distributed multi-node training
Want unified interface for 20+ cloud providers
Need to avoid vendor lock-in
Multi-cloud : AWS, GCP, Azure, Kubernetes, Lambda, RunPod, 20+ providersCost optimization : Automatic cheapest cloud/region selectionSpot instances : 3-6x cost savings with automatic recoveryDistributed training : Multi-node jobs with gang schedulingManaged jobs : Auto-recovery, checkpointing, fault toleranceSky Serve : Model serving with autoscaling Use alternatives instead:
Modal : For simpler serverless GPU with Python-native APIRunPod : For single-cloud persistent podsKubernetes : For existing K8s infrastructureRay : For pure Ray-based orchestration
Quick start
Installation pip install "skypilot[aws,gcp,azure,kubernetes]"
sky check
Hello World resources:
accelerators: T4:1
run: |
nvidia-smi
echo "Hello from SkyPilot!"
sky launch -c hello hello.yaml
ssh hello
sky down hello
Core concepts
Task YAML structure
name: my-task
resources:
cloud: aws
region: us-west-2
accelerators: A100:4
cpus: 8 +
memory: 32 +
use_spot: true
disk_size: 256
num_nodes: 2
workdir: .
setup: |
pip install -r requirements.txt
run: |
python train.py
Key commands Command Purpose sky launchLaunch cluster and run task sky execRun task on existing cluster sky statusShow cluster status sky stopStop cluster (preserve state) sky downTerminate cluster sky logsView task logs sky queueShow job queue sky jobs launchLaunch managed job sky serve upDeploy serving endpoint
GPU configuration
Available accelerators
accelerators: T4:1
accelerators: L4:1
accelerators: A10G:1
accelerators: L40S:1
accelerators: A100:4
accelerators: A100-80GB:8
accelerators: H100:8
accelerators: V100:4
accelerators: TPU-v4-8
GPU fallbacks resources:
accelerators:
H100: 8
A100-80GB: 8
A100: 8
any_of:
- cloud: gcp
- cloud: aws
- cloud: azure
Spot instances resources:
accelerators: A100:8
use_spot: true
spot_recovery: FAILOVER
Cluster management
Launch and execute
sky launch -c mycluster task.yaml
sky exec mycluster another_task.yaml
ssh mycluster
sky logs mycluster
Autostop resources:
accelerators: A100:4
autostop:
idle_minutes: 30
down: true
sky autostop mycluster -i 30 --down
Cluster status
Distributed training
Multi-node setup resources:
accelerators: A100:8
num_nodes: 4
setup: |
pip install torch torchvision
run: |
torchrun \
--nnodes=$SKYPILOT_NUM_NODES \
--nproc_per_node=$SKYPILOT_NUM_GPUS_PER_NODE \
--node_rank=$SKYPILOT_NODE_RANK \
--master_addr=$(echo "$SKYPILOT_NODE_IPS" | head -n1) \
--master_port=12355 \
train.py
Environment variables Variable Description SKYPILOT_NODE_RANKNode index (0 to num_nodes-1) SKYPILOT_NODE_IPSNewline-separated IP addresses SKYPILOT_NUM_NODESTotal number of nodes SKYPILOT_NUM_GPUS_PER_NODEGPUs per node
Head-node-only execution run: |
if [ "${SKYPILOT_NODE_RANK} " == "0" ]; then
python orchestrate.py
fi
Managed jobs
Spot recovery
sky jobs launch -n my-job train.yaml
Checkpointing name: training-job
file_mounts:
/checkpoints:
name: my-checkpoints
store: s3
mode: MOUNT
resources:
accelerators: A100:8
use_spot: true
run: |
python train.py \
--checkpoint-dir /checkpoints \
--resume-from-latest
Job management
sky jobs queue
sky jobs logs my-job
sky jobs cancel my-job
File mounts and storage
Local file sync workdir: ./my-project
file_mounts:
/data/config.yaml: ./config.yaml
~/.vimrc: ~/.vimrc
Cloud storage file_mounts:
/datasets:
source: s3://my-bucket/datasets
mode: MOUNT
/models:
source: gs://my-bucket/models
mode: COPY
/outputs:
name: my-outputs
store: s3
mode: MOUNT_CACHED
Storage modes Mode Description Best For MOUNTStream from cloud Large datasets, read-heavy COPYPre-fetch to disk Small files, random access MOUNT_CACHEDCache with async upload Checkpoints, outputs
Sky Serve (Model Serving)
Basic service
service:
readiness_probe: /health
replica_policy:
min_replicas: 1
max_replicas: 10
target_qps_per_replica: 2.0
resources:
accelerators: A100:1
run: |
python -m vllm.entrypoints.openai.api_server \
--model meta-llama/Llama-2-7b-chat-hf \
--port 8000
sky serve up -n my-service service.yaml
sky serve status
sky serve status my-service
Autoscaling policies service:
replica_policy:
min_replicas: 1
max_replicas: 10
target_qps_per_replica: 2.0
upscale_delay_seconds: 60
downscale_delay_seconds: 300
load_balancing_policy: round_robin
Cost optimization
Automatic cloud selection
resources:
accelerators: A100:8
sky launch task.yaml --dryrun
Cloud preferences resources:
accelerators: A100:8
any_of:
- cloud: gcp
region: us-central1
- cloud: aws
region: us-east-1
- cloud: azure
Environment variables envs:
HF_TOKEN: $HF_TOKEN
WANDB_API_KEY: $WANDB_API_KEY
secrets:
- HF_TOKEN
- WANDB_API_KEY
Common workflows
Workflow 1: Fine-tuning with checkpoints name: llm-finetune
file_mounts:
/checkpoints:
name: finetune-checkpoints
store: s3
mode: MOUNT_CACHED
resources:
accelerators: A100:8
use_spot: true
setup: |
pip install transformers accelerate
run: |
python train.py \
--checkpoint-dir /checkpoints \
--resume
Workflow 2: Hyperparameter sweep name: hp-sweep-${RUN_ID}
envs:
RUN_ID: 0
LEARNING_RATE: 1e-4
BATCH_SIZE: 32
resources:
accelerators: A100:1
use_spot: true
run: |
python train.py \
--lr $LEARNING_RATE \
--batch-size $BATCH_SIZE \
--run-id $RUN_ID
for i in {1..10}; do
sky jobs launch sweep.yaml \
--env RUN_ID=$i \
--env LEARNING_RATE=$(python -c "import random; print(10**random.uniform(-5,-3))" )
done
Debugging
ssh mycluster
sky logs mycluster
sky queue mycluster
sky jobs logs my-job
Common issues Issue Solution Quota exceeded Request quota increase, try different region Spot preemption Use sky jobs launch for auto-recovery Slow file sync Use MOUNT_CACHED mode for outputs GPU not available Use any_of for fallback clouds
References
Resources 