Loading...
Loading...
Found 402 Skills
Use when launching cloud VMs, Kubernetes pods, or Slurm jobs for GPU/TPU/CPU workloads, training or fine-tuning models on cloud GPUs, deploying inference servers (vllm, TGI, etc.) with autoscaling, writing or debugging SkyPilot task YAML files, using spot/preemptible instances for cost savings, comparing GPU prices across clouds, managing compute across 25+ clouds, Kubernetes, Slurm, and on-prem clusters with failover between them, troubleshooting resource availability or SkyPilot errors, or optimizing cost and GPU availability.
This skill should be used when the user asks to "build a 3D hero section or WebGL background", "animate a Three.js scene", "play or crossfade GLTF animation clips", "move/lerp the camera or link it to scroll", "render thousands of objects with InstancedMesh", "set up React Three Fiber (R3F) motion", "add bloom/DOF post-processing", or "keep a Three.js/R3F animation smooth and leak-free (dispose geometries/materials/textures, free GPU memory on unmount)". Covers scenes, render loops, camera moves, GLTF AnimationMixer, instancing, scroll-linked 3D, R3F, and animation performance/disposal.
Guidance for implementing tensor parallelism in PyTorch, including ColumnParallelLinear and RowParallelLinear layers. This skill should be used when implementing distributed tensor parallel operations, sharding linear layers across multiple GPUs, or simulating collective operations like all-gather and all-reduce for parallel computation.
Use when "Modal", "serverless GPU", "cloud GPU", "deploy ML model", or asking about "serverless containers", "GPU compute", "batch processing", "scheduled jobs", "autoscaling ML"
Deploy vLLM using Docker (pre-built images or build-from-source) with NVIDIA GPU support and run the OpenAI-compatible server.
How to use the gpui-component UI library in GPUI applications, and the normative Design and Coding Guides that govern it. Use when building UIs with gpui-component components (Button, Input, Select, Dialog, Tabs, Sidebar, List, Table, etc.), setting up the library, handling component state or theming, finding the right component for a UI need, and also when designing layouts, spacing, visual hierarchy, or interaction states, writing interface copy, or making application architecture, state-ownership, or public API decisions.
GPUI styling system including theme design, responsive layouts, visual design patterns, and style composition. Use when user needs help with styling, theming, or visual design in GPUI.
Pueue job queue for long-running tasks on remote GPU workstations. TRIGGERS - run on bigblack, run on littleblack, queue job, long-running task, cache population, batch processing, GPU workstation.
Manage cloud development sandboxes with cloudrouter. Create, sync, and access remote VMs with GPU support, Docker, and browser automation. Use when asked to create a sandbox, spin up a dev environment, run code in the cloud, use GPUs, automate a browser, or interact with remote VMs.
Use this skill when profiling or optimizing a PixiJS v8 app for FPS, draw calls, or GPU memory. Covers destroy patterns (cacheAsTexture(false), releaseGlobalResources), GCSystem and TextureGCSystem, PrepareSystem, object pooling, batching rules, BitmapText for dynamic text, culling (Culler, CullerPlugin, cullable, cullArea), resolution/antialias tradeoffs. Triggers on: FPS, jank, draw calls, batching, object pool, GCSystem, PrepareSystem, Culler, cacheAsTexture, memory leak, destroy patterns.
Battle-tested PyTorch training recipes for all domains — LLMs, vision, diffusion, medical imaging, protein/drug discovery, spatial omics, genomics. Covers training loops, optimizer selection (AdamW, Muon), LR scheduling, mixed precision, debugging, and systematic experimentation. Use when training or fine-tuning neural networks, debugging loss spikes or OOM, choosing architectures, or optimizing GPU throughput.
Fine-tune any HuggingFace CV / VLM / LLM model on local NVIDIA GPUs inside an NGC PyTorch container. Use when the user wants to fine-tune a HuggingFace model (full or LoRA), train a vision / VLM / LLM model end-to-end, generate a reproducible HF training pipeline, smoke-test a HuggingFace model locally before scale-up, push a fine-tuned model to the HF Hub with a model card, or emit a self-contained rerun skill for an existing HuggingFace finetune. Supports image classification, object detection, semantic / instance / panoptic segmentation, depth estimation, image-text-to-text VLM (SFT / LoRA), and LLM SFT / DPO / GRPO. Six-step workflow: inspect and qualify, hardware and NGC image, research, generate and smoke, train + eval + infer, push and emit rerun skill.