Loading...
Loading...
Found 418 Skills
Add a new cuTile GPU kernel operator to TileGym. Covers dispatch registration in ops.py, cuTile backend implementation, __init__.py exports, test creation, and benchmark in tests/benchmark. Use when adding, creating, or implementing a new cuTile operator/kernel in TileGym, or when asking how to register a new cuTile op.
Installs or verifies AMD Quark and ensures the selected Python environment has an accelerator-matched PyTorch. Applies to "install Quark", "set up Quark", "pip install amd-quark", Quark dependency or import failures, ModuleNotFoundError for quark, and Quark kernel compiler errors. Defaults to GPU and requires explicit confirmation before CPU mode. Covers PyPI universal, exactly matched native, local-wheel, and local-source installs. Does not handle standalone PyTorch requests, ONNX Runtime-only setup, or quantization.
Implementing custom elements using GPUI's low-level Element API (vs. high-level Render/RenderOnce APIs). Use when you need maximum control over layout, prepaint, and paint phases for complex, performance-critical custom UI components that cannot be achieved with Render/RenderOnce traits.
Visualize a specific transformer decoder layer from an AutoDeploy FX graph text dump as a hierarchical DOT/PNG diagram. Optionally annotate nodes with actual GPU kernel names and durations from an nsys trace. Use when the user wants to visualize, inspect, or debug a layer in an AutoDeploy model graph dump. Triggers on: "visualize layer", "show layer", "graph of layer", "layer visualization", "dump graph layer". Assumes graph dumps already exist in a directory (produced by AD_DUMP_GRAPHS_DIR).
Build Three.js particle systems, point clouds, large repeated datasets, and CPU or GPU simulations.
Focus management and keyboard navigation in GPUI. Use when handling focus, focus handles, or keyboard navigation. Enables keyboard-driven interfaces with proper focus tracking and navigation between focusable elements.
Create new GPUI components. Use when building components, writing UI elements, or creating new component implementations.
Use when working with ANY GPU rendering, Metal, OpenGL migration, shaders, 3D content, RealityKit, AR, or display performance. Covers Metal migration, shader conversion, RealityKit ECS, RealityView, variable refresh rate, ProMotion.
A Just-In-Time (JIT) compiler for Python that translates a subset of Python and NumPy code into fast machine code. Developed by Anaconda, Inc. Highly effective for accelerating loops, custom mathematical functions, and complex numerical algorithms. Use for @njit, @vectorize, prange, cuda.jit, numba.typed, JIT compilation, parallel loops, GPU acceleration with CUDA, Monte Carlo simulations, numerical algorithms, and high-performance Python computing.
Structure prediction using Boltz-1/Boltz-2, an open biomolecular structure predictor. Use this skill when: (1) Predicting protein complex structures, (2) Validating designed binders, (3) Need open-source alternative to AF2, (4) Predicting protein-ligand complexes, (5) Using local GPU resources. For QC thresholds, use protein-qc. For AlphaFold2 prediction, use alphafold. For Chai prediction, use chai.
This skill should be used when users want to run any workload on Hugging Face Jobs infrastructure. Covers UV scripts, Docker-based jobs, hardware selection, cost estimation, authentication with tokens, secrets management, timeout configuration, and result persistence. Designed for general-purpose compute workloads including data processing, inference, experiments, batch jobs, and any Python-based tasks. Should be invoked for tasks involving cloud compute, GPU workloads, or when users mention running jobs on Hugging Face infrastructure without local setup.
Techniques for reducing peak GPU memory in Megatron Bridge — expandable segments, parallelism resizing, activation recompute, CPU offloading constraints, and common OOM fixes.