Load a sharded, on-disk dataset (sharded .npy, Parquet/Arrow, raw binary, sharded HDF5, custom layouts) into a distributed cuPyNumeric ndarray via a manual partition + leaf @task launch with CPU/OMP/GPU variants.
nvidia-cupynumeric-parallel-data-load/SKILL.md
---
name: cupynumeric-parallel-data-load
description: Load a sharded, on-disk dataset (sharded .npy, Parquet/Arrow, raw binary, sharded HDF5, custom layouts) into a distributed cuPyNumeric ndarray via a manual partition + leaf @task launch with CPU/OMP/GPU variants. Use when no single-call loader fits, including when per-shard row counts differ across files. Prefer cupynumeric.load or legate.io.hdf5.from_file when they apply.
license: CC-BY-4.0 OR Apache-2.0
compatibility: linux-x86_64, linux-aarch64, darwin-aarch64, wsl-x86_64
metadata:
version: "1.0.0"
author: "NVIDIA Corporation <legate@nvidia.com>"
upstream: https://github.com/nv-legate/cupynumeric
docs: https://docs.nvidia.com/cupynumeric/latest/
tags:
- cupynumeric
- legate
- data-loading
- io
- distributed
- parallel
- gpu
- sharded-data
---
# Parallel sharded data -> cupynumeric load
**Why this skill exists.** cupynumeric mirrors NumPy's array API,
including `cupynumeric.load` for a single `.npy` file. Beyond that,… load the full skill through Skill MeIn any connected conversation, say:
Install the Cupynumeric Parallel Data Load skill
If full content is available, it applies in this conversation and stays installed for future sessions.
Not connected yet? It takes about a minute.
Connect your AIOr add it to a project with the CLI
npx @skillme/cli add nvidia-cupynumeric-parallel-data-loadMCP endpoint
https://skillme.dev/api/mcpUse it across a team
Pin this skill in a team library and every member's AI loads this exact version until an admin updates it.
How Teams works