How does it handle hardware selection?

It uses a parameter-based guide to recommend hardware, ranging from basic CPU for Inference API calls to A100-large for models exceeding 30B parameters.

Can I deploy LoRA adapters with this skill?

Absolutely. It automatically identifies the base model from the adapter configuration and generates the required PEFT loading logic for the deployment.

Does this skill support ZeroGPU?

Yes, it includes specialized logic to configure Spaces for ZeroGPU (zero-a10g), providing free on-demand GPU access with the necessary @spaces.GPU decorators.

Which SDKs are supported?

The skill primarily focuses on Gradio for interactive ML apps, but also provides support for Streamlit and Docker SDK configurations.

Does it manage the Hugging Face CLI?

Yes, it provides the correct syntax for creating repositories, uploading files, and inspecting remote model weights directly through the HF CLI.

Hugging Face Space Deployer

Name: Hugging Face Space Deployer
Author: GhostScientist

byGhostScientist

•

Ciencia de Datos y ML

Deploys machine learning models to Hugging Face Spaces using optimized configurations for Gradio, ZeroGPU, and LoRA adapters.

Acerca de

The Hugging Face Space Deployer skill streamlines the transition from model training to public demo by automating the creation and configuration of Hugging Face Spaces. It features an intelligent decision tree that distinguishes between full models and LoRA adapters, determines Inference API compatibility, and recommends optimal hardware ranging from free ZeroGPU instances to high-performance A100s. By providing production-ready templates and managing complex dependencies like PEFT and Accelerate, this skill ensures AI engineers can showcase their work with minimal infrastructure overhead.

Características Principales

Pre-built Gradio templates for chat, image generation, and model comparison
Automatic model type detection for full weights vs. LoRA adapters
ZeroGPU optimization for free on-demand GPU inference
1 GitHub stars
Hardware recommendation engine based on parameter count and budget
Seamless integration with Hugging Face Hub CLI for repo management

Casos de Uso

Deploying LoRA adapters by automatically linking the correct base model
Launching a public chat interface for a newly fine-tuned LLM
Scaling ML prototypes using cost-effective Inference API or ZeroGPU hosting

Acerca de

Características Principales

Pre-built Gradio templates for chat, image generation, and model comparison
Automatic model type detection for full weights vs. LoRA adapters
ZeroGPU optimization for free on-demand GPU inference
1 GitHub stars
Hardware recommendation engine based on parameter count and budget
Seamless integration with Hugging Face Hub CLI for repo management

Casos de Uso

Deploying LoRA adapters by automatically linking the correct base model
Launching a public chat interface for a newly fine-tuned LLM
Scaling ML prototypes using cost-effective Inference API or ZeroGPU hosting