Upload folder using huggingface_hub

README.md

@@ -0,0 +1,44 @@
+---
+license: apache-2.0
+tags:
+- geolocation
+- vision
+- siglip
+- clip
+- geoclip
+datasets:
+- osv5m
+pipeline_tag: image-feature-extraction
+---
+
+# 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
+
+```python
+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)
+```

config.json

@@ -0,0 +1,22 @@
+{
+  "model_type": "geoclip",
+  "encoder_name": "siglip2",
+  "backbone": "google/siglip2-so400m-patch16-512",
+  "image_resolution": 512,
+  "embedding_dim": 512,
+  "location_encoder": {
+    "sigma": [
+      1,
+      16,
+      256
+    ],
+    "hidden_dim": 1024
+  },
+  "queue_size": 15360,
+  "training": {
+    "steps": 206000,
+    "batch_size": 256,
+    "lr": 0.0003,
+    "precision": "bf16"
+  }
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d859b188a68aee3ca8417dea77f34e52ef03df26dc89e6ab0e243216b86b00ac
+size 903113660