hugging-face-jobs

Run CPU/GPU workloads on Hugging Face Jobs with Docker or UV scripts.

Updated Mar 18, 2026
One-click install
npx skills add https://github.com/AnushkaTandon26/AI-ML-Projects --skill hugging-face-jobs-anushkatandon26
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: hugging-face-jobs
Source: https://github.com/AnushkaTandon26/AI-ML-Projects/tree/main/hugging-face-skills/skills/hugging-face-jobs
Command: npx skills add https://github.com/AnushkaTandon26/AI-ML-Projects --skill hugging-face-jobs-anushkatandon26

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires datasets, flashinfer-python, huggingface-hub[hf_transfer], hf-xet>= 1.1.7, torch, transformers, vllm>=0.8.5, vllm>=0.6.5, numpy, tqdm, scikit-learn, polars>=1.31.0, ascii-graph, and includes scripts (resource) and references (resource) components.

What problem does it solve?

Hugging Face Jobs enables running CPU/GPU workloads in the cloud without local setup, streamlining scalable experimentation and batch processing.

Core Features & Use Cases

  • UV scripts and Docker-based jobs to execute Python workloads on hosted infrastructure.
  • Hardware selection, cost estimation, authentication with tokens, secrets management, and timeout configuration.
  • Result persistence by pushing outputs to the Hugging Face Hub or external storage, enabling reproducibility and sharing.
  • Use cases include data processing, batch inference, model evaluation, and long-running experiments.

Quick Start

Submit a Hugging Face Jobs task using a secret HF_TOKEN, select an appropriate hardware flavor, and push results back to the Hub for persistence.

Frequently Asked Questions about hugging-face-jobs

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run batch inference workloads on Hugging Face cloud compute securely?

To run batch inference on Hugging Face cloud compute, submit a workload using UV scripts or Docker-based jobs, authenticate with a secret HF_TOKEN, select a hardware flavor, and push outputs to the Hub for persistence.

Can I use Docker to run long-running experiments on Hugging Face Jobs?

Yes, Hugging Face Jobs supports Docker-based jobs to execute long-running experiments on hosted infrastructure, allowing configurable timeouts and hardware selection to streamline scalable experimentation.

How do I manage secrets and authentication tokens for Hugging Face Jobs tasks?

Hugging Face Jobs enforces secure token handling by passing your HF_TOKEN as a secret during job submission, ensuring authentication credentials are protected throughout the cloud compute workload execution.

What is the best way to persist results from Hugging Face Jobs back to the Hub?

The best way to persist results from Hugging Face Jobs is pushing outputs directly to the Hugging Face Hub or external storage, which enables reproducibility and sharing of data processing and batch inference results.

Does Hugging Face Jobs support GPU hardware selection and cost estimation?

Hugging Face Jobs supports configurable GPU and CPU hardware flavors with built-in cost estimation, allowing you to select appropriate compute resources for data processing and model evaluation tasks.

Why do I need UV scripts for running Python workloads on Hugging Face infrastructure?

UV scripts provide a lightweight method to execute Python workloads on Hugging Face infrastructure without local setup, complementing Docker-based jobs for scalable batch processing and experimentation.