Downloads · 30 days
14
11% of all-time downloads
geospot-base
sdan/geospot-base
geospot-base is a image feature extraction model from sdan. Use it for the image feature extraction task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.
A geolocation model built on SigLIP2-so400m (512px) that predicts GPS coordinates from images.
All-time downloads
131
Public
Parameters
440M
903 MB on disk
Likes
2
Public
Repo makeup
Click a slice to open those files.
.safetensors903 MB · 100%
Parameter types
How the weights are stored.
BF16429M · 97%
Datasets
Model card
From the Hugging Face model README
GeoSpot Base
A geolocation model built on SigLIP2-so400m (512px) that predicts GPS coordinates from images.
Model Details
- Backbone: google/siglip2-so400m-patch16-512 (frozen)
- Image Resolution: 512x512
- Embedding Dim: 512
- Training Steps: 206k
- Training Data: ~10.6M streetview images
Architecture
GeoCLIP-style contrastive learning between:
- Image Encoder: SigLIP2 vision tower + MLP projection (1152 → 512)
- Location Encoder: Multi-scale RFF encoding with learnable capsules
Usage
from geoclip.model.GeoCLIP import GeoCLIP
import torch
model = GeoCLIP(from_pretrained=False, encoder_name="siglip2")
state_dict = torch.load("model.safetensors")
model.load_state_dict(state_dict)
# Predict location from image
top_gps, top_probs = model.predict("image.jpg", top_k=5)
- Task
- Image Feature Extraction
- Type
- geoclip
- License
- apache-2.0
- Created
- Dec 6, 2025
- Updated
- Dec 6, 2025