Download

Skill UI

Browse and discover 5987+ curated skills

All Development Artificial Intelligence Design & Creative Product & Business Data Science Marketing Soft Skills Productivity Engineering Languages

Search Ray , found 14 results

Default Newest Most Downloaded

OpenRLHF High Performance Training

openrlhf-training

Orchestra-Research/AI-Research-SKILLs

OpenRLHF is a Ray-based RLHF framework that accelerates large-model PPO, GRPO, RLOO, and DPO training with vLLM inference and ZeRO-3 resource sharing, targeting distributed GPU clusters handling 7B-70B+ parameter models.

Ray Data Processing

Orchestra-Research/AI-Research-SKILLs

Ray Data is a scalable distributed data processing library for ML workloads, streaming execution across CPU/GPU, integrates with Ray Train, PyTorch, TensorFlow, and handles Parquet/CSV/JSON/images for preprocessing, ETL, or batch inference from single node to cluster.

Ray Train Distributed Orchestration

Orchestra-Research/AI-Research-SKILLs

Ray Train orchestrates distributed PyTorch/TensorFlow/HuggingFace training from laptops to thousands of nodes, adding elastic scaling, hyperparameter search, and fault tolerance while keeping most code unchanged.

Sentry Migration Deep Dive

sentry-migration-deep-dive

jeremylongshore/claude-code-plugins-plus-skills

Guides teams through executing a full migration from Rollbar, Bugsnag, Raygun, or other error-tracking solutions into Sentry, outlining prerequisites, parallel runs, alerts, and historical data export.

Distributed Dask Workflows

K-Dense-AI/claude-scientific-skills

Dask enables larger-than-memory pandas/NumPy workflows through parallel, distributed execution across dataframes, arrays, bags, and futures, making it ideal for scaling ETL, analytics, or custom computations across cores or clusters.

FlowIO FCS Handler

K-Dense-AI/claude-scientific-skills

Lightweight Python library to read, inspect, and write FCS v2.0-3.1 files, outputting metadata, channel info, and NumPy arrays for preprocessing cytometry pipelines and backend services.

TileDB VCF Storage

K-Dense-AI/claude-scientific-skills

TileDB-VCF lets you ingest, store, query, and export large genomic VCF/BCF datasets using TileDB’s sparse arrays, supporting incremental sample growth, parallel regions queries, and cloud-local interoperability for population genomics workflows.

Zarr Python Arrays

K-Dense-AI/claude-scientific-skills

Handles chunked, compressed N-D arrays in Python for cloud-friendly workflows; integrates seamlessly with NumPy, Dask, and Xarray to enable parallel I/O, sharding, and persistence to S3/GCS for large-scale scientific pipelines.

Functional Data Transformations

fp-data-transforms

sickn33/antigravity-awesome-skills

This skill demonstrates everyday data transformations in TypeScript via functional patterns, covering arrays, object reshaping, grouping, aggregation, and null-safe access while comparing imperative and functional approaches to guide practical choices.

FP Pipe Flow Guide

sickn33/antigravity-awesome-skills

Quick fp-ts reference outlining how to use pipe for transforming specific values and flow for creating reusable transformation pipelines, with examples covering Option and Array utilities for chained data processing.

ThreeJS Interaction Guide

threejs-interaction

sickn33/antigravity-awesome-skills

Practical reference covering Three.js raycasting, input handling, and camera controls for selecting objects and building interactive 3D scenes with mouse, touch, and orbital/fly navigation.

Refactoring UI Principles

jeremylongshore/claude-code-plugins-plus-skills

Audit and refine web interfaces by enforcing grayscale-first hierarchy, consistent spacing scales, typography rules, and component styling so front-end code feels professional before adding color.

Language