Hub

Download models from HuggingFace and manage your local model library

Hub (04)

Download models from HuggingFace and manage a local library. Search by repository ID, download individual files or entire repos, and import existing local models.

Views

Search

Browse and download from HuggingFace:

1. Enter a HuggingFace repo ID (e.g., TheBloke/Mistral-7B-GGUF)
2. Click FETCH to list all files
3. Each file shows: filename, size, format badge (GGUF, SafeTensors, other)
4. Toggle Models Only to filter to model files
5. Click DOWNLOAD on individual files or DOWNLOAD REPO for everything

Progress panel shows real-time status: file name, bytes, percentage. Downloads can be cancelled.

Library

Manage locally downloaded models:

• Each model card: name, format badge, size, source repo, download date
• LOAD button to load into ForgeAI
• DELETE button to remove
• Storage summary: total models and disk usage

Import models already on disk:

• Import GGUF file
• Import SafeTensors file
• Import model folder

Hero Panel

Field	Description
MODELS	Count of local models
STORAGE	Total disk usage
SOURCE	HUGGINGFACE
FORMATS	GGUF, SAFETENSORS

Workflow: Download and Use

1. Search

   Enter repo ID (e.g., TheBloke/Mistral-7B-Instruct-v0.2-GGUF) and click FETCH

2. Download

   Click DOWNLOAD on a specific file (e.g., mistral-7b-instruct-v0.2.Q4_K_M.gguf)

3. Load

   Switch to Library view, click LOAD on the downloaded model

4. Use

   Model is now available across all modules

Downloaded models from the Hub are automatically available in the Test module’s local library chips and in the Convert module’s repository selector.