{"id":48220322,"url":"https://github.com/gpu-cli/gpu","last_synced_at":"2026-04-04T19:09:29.722Z","repository":{"id":329027463,"uuid":"1086975315","full_name":"gpu-cli/gpu","owner":"gpu-cli","description":"Public facing GPU cli docs and issues ","archived":false,"fork":false,"pushed_at":"2026-04-03T17:37:47.000Z","size":3557,"stargazers_count":5,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-04-03T20:24:34.340Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/gpu-cli.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2025-10-31T07:24:06.000Z","updated_at":"2026-04-03T17:37:51.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/gpu-cli/gpu","commit_stats":null,"previous_names":["gpu-cli/gpu"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/gpu-cli/gpu","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/gpu-cli%2Fgpu","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/gpu-cli%2Fgpu/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/gpu-cli%2Fgpu/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/gpu-cli%2Fgpu/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/gpu-cli","download_url":"https://codeload.github.com/gpu-cli/gpu/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/gpu-cli%2Fgpu/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":31409473,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-04T10:20:44.708Z","status":"ssl_error","status_checked_at":"2026-04-04T10:20:06.846Z","response_time":60,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2026-04-04T19:09:29.214Z","updated_at":"2026-04-04T19:09:29.710Z","avatar_url":"https://github.com/gpu-cli.png","language":"Python","readme":"# GPU CLI\n\nRun any code on cloud GPUs with a single command. Just prefix your normal commands with `gpu run`.\n\n```bash\npython train.py           # local\ngpu run python train.py   # remote GPU\n```\n\n## Features\n\n- **Simple** - Prefix commands with `gpu run`, that's it\n- **Fast** - Connection pooling, delta sync, real-time output streaming\n- **Cost-efficient** - Auto-stops pods when idle (save 60-98% on GPU costs)\n- **Multi-cloud** - RunPod, Vast.ai, local Docker\n- **Secure** - Zero-trust encryption on supported providers\n- **Teams** — Organizations with pooled sessions, sub-accounts, and CI/CD service tokens (Team \u0026 Enterprise)\n\n## Quick Start\n\n```bash\n# 1. Install GPU CLI\ncurl -fsSL https://gpu-cli.sh/install.sh | sh\n\n# 2. Run your code on a remote GPU\ngpu run python train.py\n```\n\n---\n\n## Claude Code Plugin\n\nThis repo includes a Claude Code plugin that supercharges GPU CLI with AI assistance. Describe what you want in plain English, and Claude generates complete, runnable GPU workflows.\n\n### What's Included\n\n#### Skills (Automatic AI Capabilities)\n\n| Skill | Description |\n|-------|-------------|\n| **gpu-workflow-creator** | Transform natural language into complete GPU projects |\n| **gpu-ml-trainer** | LLM fine-tuning, LoRA training, classifier training |\n| **gpu-inference-server** | Set up vLLM, TGI, or custom inference APIs |\n| **gpu-media-processor** | Whisper transcription, voice cloning, video generation |\n| **gpu-cost-optimizer** | GPU selection advice and cost optimization |\n| **gpu-debugger** | Debug failed runs, OOM errors, connectivity issues |\n\n#### Slash Commands\n\n| Command | Description |\n|---------|-------------|\n| `/gpu-cli:gpu-create` | Create a complete project from a description |\n| `/gpu-cli:gpu-optimize` | Analyze and optimize your gpu.jsonc |\n| `/gpu-cli:gpu-debug` | Debug a failed GPU run |\n| `/gpu-cli:gpu-quick` | Quick-start common workflows |\n\n### Example Conversations\n\n**Create a LoRA training project:**\n```\nYou: I want to train a LoRA on photos of my dog so I can generate images of it\n\nClaude: [Generates complete project with gpu.jsonc, train.py, requirements.txt, README.md]\n```\n\n**Set up a private LLM API:**\n```\nYou: Set up Llama 3.1 70B as a private ChatGPT-like API\n\nClaude: [Generates vLLM server config with OpenAI-compatible endpoints]\n```\n\n**Debug an error:**\n```\nYou: /gpu-cli:gpu-debug CUDA out of memory when running FLUX\n\nClaude: [Analyzes error, suggests reducing batch size or upgrading to A100]\n```\n\n**Optimize costs:**\n```\nYou: /gpu-cli:gpu-optimize\n\nClaude: [Reviews gpu.jsonc, suggests RTX 4090 instead of A100 for your workload, saving 75%]\n```\n\n---\n\n## Templates\n\nReady-to-use templates for common AI/ML workflows:\n\n| Template | GPU | Description |\n|----------|-----|-------------|\n| [Ollama Models](./templates/ollama-models/) | RTX 4090 | Run LLMs with Ollama - includes Web UI and OpenAI-compatible API |\n| [vLLM Models](./templates/vllm-models/) | RTX 4090 | High-performance LLM inference with vLLM |\n| [Background Removal](./templates/background-removal/) | RTX 4090 | Remove backgrounds from images using AI |\n| [CrewAI Stock Analysis](./templates/crewai-stock-analysis/) | RTX 4090 | Multi-agent stock analysis with CrewAI + Ollama |\n| [HuggingFace Gradio](./templates/huggingface-gradio/) | RTX 4090 | Run HuggingFace models with Gradio UI |\n| [Qwen Image Edit](./templates/qwen-image-edit/) | RTX 4090 | Edit images using Qwen vision model |\n\n## Common Commands\n\n```bash\n# Run a command on remote GPU\ngpu run python script.py\n\n# Run a server with port forwarding\ngpu run -p 8188:8188 python server.py --listen 0.0.0.0\n\n# Open a shell on the remote pod\ngpu shell\n\n# View running pods\ngpu pods list\n\n# Stop a pod\ngpu stop\n\n# Interactive dashboard\ngpu dashboard\n```\n\n### Team Management\n\n```bash\n# Create an organization\ngpu org create \"My Team\"\n\n# Switch to org context\ngpu org switch my-team\n\n# Invite a teammate\ngpu org invite alice@example.com --role admin\n\n# Create a CI/CD service account\ngpu org service-account create --name \"github-actions\"\n```\n\n## Configuration\n\nCreate a `gpu.jsonc` in your project:\n\n```jsonc\n{\n  \"$schema\": \"https://gpu-cli.sh/schema/v1/gpu.json\",\n  \"project_id\": \"my-project\",\n  \"provider\": \"runpod\",\n\n  // Sync outputs back to local machine\n  \"outputs\": [\"output/\", \"models/\"],\n\n  // GPU selection\n  \"gpu_type\": \"RTX 4090\",\n  \"min_vram\": 24,\n\n  // Optional: Pre-download models\n  \"download\": [\n    { \"strategy\": \"hf\", \"source\": \"black-forest-labs/FLUX.1-dev\", \"allow\": \"*.safetensors\" }\n  ],\n\n  \"environment\": {\n    \"base_image\": \"runpod/pytorch:2.4.0-py3.11-cuda12.4.1-devel-ubuntu22.04\"\n  }\n}\n```\n\n### Network Volumes (Recommended)\n\nFor faster startup and persistent model storage, use RunPod Network Volumes. See the [Network Volumes Guide](./docs/network-volumes.md) for setup instructions.\n\n## GPU Options\n\n| GPU | VRAM | Best For | Cost/hr |\n|-----|------|----------|---------|\n| RTX 4090 | 24GB | Image generation, LoRA training | ~$0.44 |\n| RTX 4080 | 16GB | SDXL, most workflows | ~$0.35 |\n| A100 40GB | 40GB | 70B models, video generation | ~$1.29 |\n| A100 80GB | 80GB | 70B+ models, large batch | ~$1.79 |\n| H100 80GB | 80GB | Maximum performance | ~$3.99 |\n\n## Documentation\n\n- [Network Volumes Guide](./docs/network-volumes.md) - Persistent storage for models\n- [Organizations Guide](https://gpu-cli.sh/docs/organizations) - Team billing, sub-accounts, and service tokens\n- [GPU CLI Docs](https://gpu-cli.sh/docs) - Full documentation\n\n## License\n\nMIT\n","funding_links":[],"categories":[],"sub_categories":[],"project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fgpu-cli%2Fgpu","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fgpu-cli%2Fgpu","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fgpu-cli%2Fgpu/lists"}