hugging-face-vision-trainer

Official

Train vision models on Hugging Face Jobs

AuthorBlackRoad-OS-Inc
Version1.0.0
Installs0

System Documentation

What problem does it solve?

Trains and fine-tunes object detection, image classification, and SAM/SAM2 vision models on Hugging Face Jobs cloud GPUs, removing the need for local GPU setup and enabling scalable experimentation.

Core Features & Use Cases

  • Supports object detection (D-FINE, RT-DETR v2, DETR, YOLOS), image classification (timm-based models, ViT, ResNet), and SAM/SAM2 segmentation with prompt-based fine-tuning.
  • Provides COCO-format dataset preparation, Albumentations-based augmentation, evaluation metrics (mAP, mAR), and Hub persistence with Trackio monitoring.
  • Includes ready-to-run training scripts, dataset validation utilities, and cost estimation to help plan hardware and budgets.

Quick Start

Run a full training job on Hugging Face Jobs for vision models by selecting a model and dataset, then push the trained model to the Hub.

Dependency Matrix

Required Modules

transformers>=5.2.0accelerate>=1.1.0albumentations>=1.4.16timmdatasets>=4.0torchmetricspycocotoolstrackiohuggingface_hubmonaitorchvisionevaluatescikit-learn

Components

scriptsreferences

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: hugging-face-vision-trainer
Download link: https://github.com/BlackRoad-OS-Inc/blackroad-operator/archive/main.zip#hugging-face-vision-trainer

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.