Hugging Face

Pack

Stack Hugging Face : Trainer, Datasets, Gradio, Spaces, Transformers.js, TRL, plus patterns PyTorch et playbooks ML — sans skills CI ni kernels.

Cette collection Arboris assemble skills et commandes autour d’un même thème. Installe le pack via la CLI ou un ZIP, puis ouvre ton agent pour enchaîner les tâches couvertes. Les éléments listés ci-dessous restent accessibles individuellement dans le catalogue si tu préfères une install à la carte. Objectif : gagner du temps et garder une base commune de prompts de qualité. Chaque skill et chaque commande peut aussi être copié seul, scanné côté sécu, et mis à jour quand le catalogue évolue. Idéal pour standardiser un workflow d’équipe sans imposer un monorepo de prompts.

13 éléments

Commandes · huggingface-starter

/commitCommit Conventional Commits + changelog/semver (message clair, scope, type)./ready-prPrépare une PR — branche, checks, changelog, commit, description MR (sans push)./review-prLance une revue de pull request structurée (risques, tests, clarté)./explain-codeExplique un fichier ou une sélection de code de façon claire et progressive.

Skills · huggingface-starter

Huggingface Datasets (Hugging Face)Use this skill for Hugging Face Dataset Viewer API workflows that fetch subset/split metadata, paginate rows, search text, apply filters, download parquet URLs, and read size or statistics.Huggingface Gradio (Hugging Face)Build Gradio web UIs and demos in Python. Use when creating or editing Gradio apps, components, event listeners, layouts, or chatbots.Huggingface Llm Trainer (Hugging Face)Train or fine-tune language and vision models using TRL (Transformer Reinforcement Learning) or Unsloth with Hugging Face Jobs infrastructure. Covers SFT, DPO, GRPO and reward modeling training methods, plus GGUF conversion for local deployment. Includes guidance on the TRL Jobs package, UV scripts with PEP 723 format, dataset preparation and validation, hardware selection, cost estimation, Trackio monitoring, Hub authentication, model selection/leaderboards and model persistence. Use for tasks involving cloud GPU training, GGUF conversion, or when users mention training on Hugging Face Jobs without local GPU setup.Huggingface Spaces (Hugging Face)Build, deploy, and maintain applications on Hugging Face Spaces — Gradio / Docker / Static SDKs, ZeroGPU and dedicated hardware, model loading, debugging, buckets, inference providers, community grants. Use whenever the user asks to create or host an app on Hugging Face, port code onto ZeroGPU, fix a Space that won't build or run, or otherwise work with `hf spaces …`, `@spaces.GPU`, Space README frontmatter, or the `spaces` Python package.ml-adoption-playbookEnd-to-end methodology for AI agents and software engineers to add machine learning algorithms to existing non-ML codebases. Covers problem framing, data readiness, architectural decoupling, and baseline model integration. Use when adding a machine learning capability to a codebase that has none, from problem framing through a baseline model.mle-workflowProduction machine-learning engineering workflow for data contracts, reproducible training, model evaluation, deployment, monitoring, and rollback. Use when building, reviewing, or hardening ML systems beyond one-off notebooks.pytorch-patternsPyTorch deep learning patterns and best practices for building robust, efficient, and reproducible training pipelines, model architectures, and data loading. Use when writing or reviewing PyTorch training loops, model architectures, or data loading, or when a run will not reproduce.Transformers JS (Hugging Face)Use Transformers.js to run state-of-the-art machine learning models directly in JavaScript/TypeScript. Supports NLP (text classification, translation, summarization), computer vision (image classification, object detection), audio (speech recognition, audio classification), and multimodal tasks. Works in browsers and server-side runtimes (Node.js, Bun, Deno) with WebGPU/WASM using pre-trained models from Hugging Face Hub.Trl Training (Hugging Face)Post-train LLMs with TRL (Transformers Reinforcement Learning) — SFT, DPO, GRPO, KTO, and reward-model training. Use when writing or debugging training code with the TRL Python API or the trl CLI.