H

Hugging Face技能

add-model-descriptions
huggingface
为HuggingFace路由器中的新模型添加描述,以更新chat-ui配置。当路由器上有新模型发布且需要描述时使用…
official
add-or-fix-type-checking
huggingface
Fixes broken typing checks detected by ty, make typing, or make check-repo. Use when typing errors appear in local runs, CI, or PR logs.
official
add-or-fix-type-checking
huggingface
修复由 ty、make typing 或 make check-repo 检测到的损坏的类型检查。当本地运行、CI 或 PR 日志中出现类型错误时使用。
official
alignment-review
huggingface
Review code changes for bugs and alignment with OpenEnv principles and RFCs. Use when reviewing PRs, checking code before commit, or when asked to review…
official
alignment-review
huggingface
审查代码变更中的错误,确保其符合OpenEnv原则和RFC规范。适用于审查PR、提交前检查代码,或被要求进行审查时。
official
cpu-kernels
huggingface
为Hugging Face内核生态系统提供使用SIMD内联函数(AVX2/AVX512)编写、优化和基准测试C++ CPU内核的指导。包括…
official
cpu-kernels
huggingface
Provides guidance for writing, optimizing, and benchmarking C++ CPU kernels with SIMD intrinsics (AVX2/AVX512) for the Hugging Face kernels ecosystem. Includes…
official
cuda-kernels
huggingface
Provides guidance for writing and benchmarking optimized CUDA kernels for NVIDIA GPUs (H100, A100, T4) targeting HuggingFace diffusers and transformers…
official
cuda-kernels
huggingface
提供针对NVIDIA GPU(H100、A100、T4)编写和基准测试优化CUDA内核的指导,目标为HuggingFace diffusers和transformers…
official
custom-blocks
huggingface
Use when the user has written (or wants to write) a `ModularPipelineBlocks` subclass in a local Python file and needs to package it into a Hub-uploadable…
official
deploy-hf
huggingface
将 OpenEnv 环境部署到 Hugging Face Spaces。当被要求部署、推送到 Hugging Face 或更新空间时使用。
official
deploy-hf
huggingface
Deploy an OpenEnv environment to Hugging Face Spaces. Use when asked to deploy, push to Hugging Face, or update a space.
official
diffusers-cli
huggingface
Use when the user wants to run a diffusers pipeline from a terminal (one-off generation, batch jobs, smoke-testing a new model), run on HF Sandbox hardware via…
official
example-skill
huggingface
用于动作冒烟测试的示例夹具技能
official
example-skill
huggingface
Example fixture skill for action smoke tests
official
generate-openenv-env
huggingface
根据具体用例生成OpenEnv环境(例如,“为textarena库生成一个环境”)。当被要求设计或实现新的…时使用。
official
generate-openenv-env
huggingface
Generate OpenEnv environments from a concrete use case (for example, "generate an env for the library textarena"). Use when asked to design or implement a new…
official
gradio
huggingface
在Python中构建Gradio网页界面和演示。用于创建或编辑Gradio应用、组件、事件监听器、布局或聊天机器人时。
official
hf-cli
huggingface
Hugging Face Hub CLI (`hf`) for downloading, uploading, and managing models, datasets, spaces, buckets, repos, papers, jobs, and more on the Hugging Face Hub.…
official
hf-cli
huggingface
Hugging Face Hub CLI (`hf`) for downloading, uploading, and managing repositories, models, datasets, and Spaces on the Hugging Face Hub. Replaces now…
official
hf-cli
huggingface
Hugging Face Hub CLI (`hf`) for downloading, uploading, and managing repositories, models, datasets, and Spaces on the Hugging Face Hub. Replaces now…
official
hf-cloud-aws-context-discovery
huggingface
在任何AWS任务开始时,发现用户的本地AWS上下文(活跃配置文件、区域、账户ID、调用者身份)。在执行其他AWS操作之前,先使用此技能…
official
hf-cloud-python-env-setup
huggingface
为SageMaker / AWS工作设置一个隔离的Python环境,使用正确的Python版本和当前的boto3。每当需要运行Python代码时,请使用此技能…
official
hf-cloud-sagemaker-deployment-planner
huggingface
规划和协调将模型部署到Amazon SageMaker AI。当用户想要在……上部署、托管、服务或公开模型时,使用此技能。
official
hf-cloud-sagemaker-iam-preflight
huggingface
在部署或训练前,确保存在可用的SageMaker执行角色。在即将创建SageMaker端点、模型、训练任务时使用此技能。
official
hf-cloud-sagemaker-production-defaults
huggingface
创建一个SageMaker端点(实时或异步),默认启用自动扩缩、CloudWatch告警和标签功能。每当要创建…时,使用此技能。
official
hf-cloud-serving-image-selection
huggingface
为SageMaker模型部署选择合适的服务容器,并查找其当前镜像URI。在即将部署模型到……时使用此技能。
official
hf-mcp
huggingface
通过MCP服务器工具使用Hugging Face Hub。搜索模型、数据集、Spaces、论文。获取仓库详情、获取文档、运行计算任务、使用Gradio…
official
hf-mcp
huggingface
Use Hugging Face Hub via MCP server tools. Search models, datasets, Spaces, papers. Get repo details, fetch documentation, run compute jobs, and use Gradio…
official
hf-mem
huggingface
Hugging Face CLI 用于估算从 Hugging Face Hub 加载 Safetensors 或 GGUF 模型权重进行推理所需的内存
official
hf-release-notes
huggingface
Generate Hugging Face Hub (huggingface_hub) release notes from cached PR JSON files. Use when asked to draft release notes from PR files.
official
hf-release-notes
huggingface
根据缓存的PR JSON文件生成Hugging Face Hub (huggingface_hub) 发布说明。当要求从PR文件起草发布说明时使用。
official
hf-space-recovery
huggingface
Diagnose and recover failing or stuck Hugging Face Space deployments for OpenEnv environments. Use when deploying envs from `envs/` to the Hub (`openenv`…
official
hf-space-recovery
huggingface
Diagnose and recover failing or stuck Hugging Face Space deployments for OpenEnv environments. Use when deploying envs from `envs/` to the Hub (`openenv`…
official
Hugging Face Cli
huggingface
Execute Hugging Face Hub operations using the `hf` CLI. Use when the user needs to download models/datasets/spaces, upload files to Hub repositories, create repos, manage local cache, or run compute jobs on HF infrastructure. Covers authentication, file transfers, repository creation, cache operations, and cloud compute.
official
Hugging Face Datasets
huggingface
在 Hugging Face Hub 上创建和管理数据集。支持初始化仓库、定义配置/系统提示、流式更新行数据,以及基于 SQL 的数据集查询与转换。设计为与 HF MCP 服务器协同工作,实现完整的数据集工作流。
official
Hugging Face Evaluation
huggingface
在Hugging Face模型卡片中添加和管理评估结果。支持从README内容中提取评估表格、从Artificial Analysis API导入分数,以及使用vLLM/lighteval运行自定义模型评估。兼容model-index元数据格式。
official
Hugging Face Jobs
huggingface
在Hugging Face Jobs基础设施上运行任意工作负载。涵盖UV脚本、基于Docker的任务、硬件选择、成本估算、令牌认证、密钥管理、超时配置及结果持久化。专为通用计算工作负载设计,包括数据处理、推理、实验、批处理任务及任何基于Python的任务。
official
Hugging Face Model Trainer
huggingface
使用Hugging Face Jobs基础设施上的TRL(Transformer强化学习)训练或微调语言模型。涵盖SFT、DPO、GRPO及奖励建模训练方法,以及用于本地部署的GGUF转换。包含数据集准备、硬件选择、成本估算和模型持久化的指导。
official
Hugging Face Paper Publisher
huggingface
在Hugging Face Hub上发布和管理研究论文。支持创建论文页面、将论文链接到模型/数据集、声明作者身份,以及生成基于Markdown的专业研究文章。
official
Hugging Face Tool Builder
huggingface
使用Hugging Face API构建可复用的脚本和工具。适用于链式调用或组合API调用,以及需要重复/自动化执行的任务。创建可复用的命令行脚本,用于从Hugging Face Hub获取、丰富或处理数据。
official
Hugging Face Trackio
huggingface
使用Trackio跟踪和可视化机器学习训练实验。在训练过程中记录指标(Python API)或检索/分析已记录的指标(CLI)时使用。支持实时仪表盘可视化、HF Space同步以及用于自动化的JSON输出。
official
hugging-face-dataset-viewer
huggingface
使用此技能处理Hugging Face Dataset Viewer API工作流,以获取子集/分割元数据、分页行、搜索文本、应用过滤器、下载parquet文件等。
official
hugging-face-evaluation
huggingface
Add evaluation results to Hugging Face model repositories using the .eval_results/ format. Uses HF CLI for PR management and manual YAML creation.
official
hugging-face-evaluation
huggingface
使用.eval_results/格式向Hugging Face模型仓库添加评估结果。通过HF CLI进行PR管理并手动创建YAML文件。
official
hugging-face-object-detection-trainer
huggingface
使用 Hugging Face Transformers 在 Hugging Face Jobs 云 GPU 上训练和微调目标检测模型(D-FINE、RT-DETR v2、DETR、YOLOS)。涵盖…
official
hugging-face-paper-pages
huggingface
以Markdown格式查找并阅读Hugging Face论文页面,并使用论文API获取结构化元数据,如作者、关联的模型/数据集/空间、GitHub等。
official
hugging-face-vision-trainer
huggingface
训练和微调视觉模型,用于目标检测(D-FINE、RT-DETR v2、DETR、YOLOS)和图像分类(timm模型——MobileNetV3、MobileViT、ResNet等)
official
huggingface-best
huggingface
通过查询官方Hugging Face基准排行榜,为任务寻找最佳模型,补充模型大小数据,筛选出适合用户设备的模型,并返回包含基准分数的对比表格。
official
huggingface-community-evals
huggingface
使用 inspect-ai 和 lighteval 在本地硬件上对 Hugging Face Hub 模型运行评估。用于后端选择、本地 GPU 评估,以及选择…
official
huggingface-datasets
huggingface
将此技能用于Hugging Face Dataset Viewer API工作流,以获取子集/分割元数据、分页行、搜索文本、应用过滤器、下载parquet…
official
huggingface-gradio
huggingface
在Python中构建Gradio网页界面和演示。适用于创建或编辑Gradio应用、组件、事件监听器、布局或聊天机器人时使用。
official
huggingface-jobs
huggingface
当用户希望在Hugging Face Jobs基础设施上运行任何工作负载时,应使用此技能。涵盖UV脚本、基于Docker的任务、硬件选择等。
official
huggingface-llm-trainer
huggingface
使用TRL(Transformer强化学习)或Unsloth,借助Hugging Face Jobs基础设施训练或微调语言与视觉模型。涵盖SFT、DPO、…
official
huggingface-local-models
huggingface
用于选择在本地通过llama.cpp和GGUF在CPU、Mac Metal、CUDA或ROCm上运行的模型。涵盖查找GGUF、量化选择、运行服务器、精确…
official
huggingface-lora-space-builder
huggingface
根据用户提供的LoRA,在Hugging Face Spaces上构建并发布一个Gradio演示。当有人要求创建、生成、发布或部署Space、演示等时使用。
official
huggingface-paper-publisher
huggingface
在 Hugging Face Hub 上发布和管理研究论文。支持创建论文页面、将论文链接到模型/数据集、声明作者身份以及生成…
official
huggingface-papers
huggingface
在Markdown中查找并阅读Hugging Face论文页面,并使用论文API获取结构化元数据,如作者、关联的模型/数据集/空间、GitHub等。
official
huggingface-spaces
huggingface
在 Hugging Face Spaces 上构建、部署和维护应用程序 — Gradio / Docker / 静态 SDK、ZeroGPU 和专用硬件、模型加载、调试……
official
huggingface-tool-builder
huggingface
当用户想要构建工具/脚本或完成某项任务,且使用Hugging Face API的数据会有所帮助时,使用此技能。这在以下情况下尤其有用…
official
huggingface-trackio
huggingface
使用Trackio跟踪和可视化机器学习训练实验。在训练期间记录指标(Python API)、触发训练诊断警报或…时使用。
official
huggingface-vision-trainer
huggingface
训练和微调视觉模型,用于目标检测(D-FINE、RT-DETR v2、DETR、YOLOS)、图像分类(timm模型——MobileNetV3、MobileViT、ResNet……)
official
huggingface-zerogpu
huggingface
AI demos and GPU compute with Gradio Spaces and Hugging Face Spaces ZeroGPU. Use when writing or reviewing code that uses `@spaces.GPU`, configuring…
official
implement
huggingface
Make tests pass. Invoke after /write-tests produces failing tests.
official
implement
huggingface
使测试通过。在/write-tests生成失败的测试后调用。
official
inference-server
huggingface
启动并测试prime-rl推理服务器。当被要求运行推理、启动vLLM、测试模型或启动推理服务器时使用。
official
init-physics-intern
huggingface
Scaffold a PhysicsIntern research workspace in the current folder. Run only when the user explicitly invokes it to set up a new physics/maths research…
official
init-physics-intern
huggingface
Scaffold a PhysicsIntern research workspace in the current folder.
official
installation
huggingface
如何安装prime-rl及其可选依赖项。在设置项目、安装如deep-gemm用于FP8模型等额外组件,或进行故障排除时使用…
official
integrating-models
huggingface
在向diffusers添加新模型或流水线时使用,用于为新模型设置文件结构、将流水线转换为模块化格式,或转换权重…
official
investigate-run
huggingface
Post-mortem analysis of a PhysicsIntern workspace run (Claude Code, Pi, Codex, or OpenCode host). Reconstructs the trajectory from the session record — JSONL…
official
investigate-run
huggingface
对PhysicsIntern工作区运行(Claude Code、Pi或Codex主机)的事后分析。从会话JSONL文件中重建轨迹,审计…
official
openenv-cli
huggingface
OpenEnv CLI (`openenv`) for scaffolding, validating, building, and pushing OpenEnv environments.
official
openenv-cli
huggingface
OpenEnv CLI (`openenv`) for scaffolding, validating, building, and pushing OpenEnv environments.
official
pre-submit-pr
huggingface
在提交拉取请求前验证更改。运行全面检查,包括代码检查、测试、对齐审查和RFC分析。在创建…之前使用。
official
pre-submit-pr
huggingface
Validate changes before submitting a pull request. Run comprehensive checks including lint, tests, alignment review, and RFC analysis. Use before creating a…
official
release
huggingface
Release workflow for deploying OpenEnv environments to Hugging Face Spaces and keeping canonical references in sync.
official
release
huggingface
将OpenEnv环境部署到Hugging Face Spaces并保持规范引用同步的发布工作流。
official
rfc-check
huggingface
Determine if proposed changes require an RFC. Use when planning significant changes, before starting major work, or when asked whether an RFC is needed.
official
rfc-check
huggingface
判断提议的变更是否需要RFC。在规划重大变更时、开始重要工作前,或被问及是否需要RFC时使用。
official
rocm-kernels
huggingface
Provides guidance for writing and benchmarking optimized Triton kernels for AMD GPUs (MI355X, R9700) on ROCm, targeting HuggingFace diffusers (LTX-Video, SD3,…
official
rocm-kernels
huggingface
提供在ROCm上为AMD GPU(MI355X、R9700)编写和基准测试优化Triton内核的指导,针对HuggingFace diffusers(LTX-Video、SD3等)
official
self-review
huggingface
Use before opening a PR, or whenever asked to self-review a diffusers contribution. Applies the same rubric as the `@claude` CI (checks the diff against…
official
simplify
huggingface
Refactor code after tests pass. The "Refactor" phase of Red-Green-Refactor.
official
simplify
huggingface
在测试通过后重构代码。红-绿-重构中的“重构”阶段。
official
sprint
huggingface
Work on a batch of GitHub issues in parallel using Agent Teams. Creates one worktree per issue with TDD enforcement, coordinates via a lead agent, then…
official
sprint
huggingface
使用Agent Teams并行处理一批GitHub issues。为每个issue创建一个worktree并强制执行TDD,通过主代理协调,然后…
official
sync-models
huggingface
Sync chat-ui's model config with the HuggingFace router — add descriptions for new models, flag reasoning-capable ones, enable artifacts for models with 32B+…
official
testing-parity
huggingface
用于调试或验证管道实现之间的数值一致性(例如,研究仓库与diffusers,标准版与模块化版)。也适用于当…
official
toml-config
huggingface
如何在prime-rl中编写和使用TOML配置。用于创建配置文件、使用配置运行命令或通过CLI覆盖配置值。
official
train-sentence-transformers
huggingface
Train or fine-tune sentence-transformers models across `SentenceTransformer` (bi-encoder; dense or static embedding model; for retrieval, similarity,…
official
transformers-js
huggingface
使用Transformers.js直接在JavaScript/TypeScript中运行最先进的机器学习模型。支持自然语言处理(文本分类、翻译、…)
official
transformers-to-mlx
huggingface
将Hugging Face transformers语言模型转换为MLX格式,用于Apple Silicon推理。当被要求将模型从transformers转换为MLX、移植……时使用。
official
trl-training
huggingface
使用TRL(Transformers强化学习)训练和微调Transformer语言模型。支持SFT、DPO、GRPO、KTO、RLOO和奖励模型训练…
official
trl-training
huggingface
Train and fine-tune transformer language models using TRL (Transformers Reinforcement Learning). Supports SFT, DPO, GRPO, KTO, RLOO and Reward Model training…
official
trl-training
huggingface
使用TRL(Transformers强化学习)训练和微调Transformer语言模型。支持SFT、DPO、GRPO、KTO、RLOO和奖励模型训练…
official
update-docs
huggingface
Update documentation across the repo after API changes. Finds stale references in docs, examples, docstrings, and fixes them.
official
update-docs
huggingface
在API变更后更新整个仓库的文档。查找文档、示例、文档字符串中的过时引用并进行修复。
official
watch-pr
huggingface
Monitor a PR's CI checks and Greptile code review after submission. Polls CI status, auto-fixes failures via ralph-loop, waits for Greptile review, addresses…
official
watch-pr
huggingface
监控PR提交后的CI检查与Greptile代码审查。轮询CI状态,通过ralph-loop自动修复失败项,等待Greptile审查,处理…
official
work-on-issue
huggingface
Start work on a GitHub issue. Extracts requirements, creates worktree, sets up TDD workflow.
official
work-on-issue
huggingface
开始处理一个GitHub issue。提取需求,创建工作树,设置TDD工作流。
official
write-tests
huggingface
Write failing tests from requirements. Invoke for each todo before /implement.
official
write-tests
huggingface
根据需求编写失败的测试。在/implement之前,为每个待办事项调用。
official
xpu-kernels
huggingface
Provides guidance for writing, optimizing, and benchmarking Triton kernels for Intel XPU GPUs (Battlemage/Arc Pro B50) using the Xe-Forge optimization…
official
xpu-kernels
huggingface
提供关于使用Xe-Forge优化框架为Intel XPU GPU(Battlemage/Arc Pro B50)编写、优化和基准测试Triton内核的指导…
official