Fetching metadata from the HF Docker repository...
-
Update torch to 2.4.0 and torchvision to 0.19.0 Running
-
Add OOD Detector logic and artifacts
-
✨ Add WhatsApp, Instagram, and LinkedIn buttons
-
✨ Add Social Share buttons to results
-
Prevent duplicate image uploads
-
Add image saving logic secure
-
Update Google Analytics config for iframes
-
Fix Dockerfile and dependencies
-
Fix libGL error by installing system libraries
-
🚀 Switch to opencv-python-headless (No system packages needed)
-
Add packages.txt to fix libGL crash
-
🔧 Add packages.txt to fix libGL error
-
🔧 Add full system dependencies for OpenCV
-
🔧 Add system dependencies (libGL) for OpenCV
-
Fix requirements typo: separate pyyaml and scikit-learn
-
🔧 Add scikit-learn and matplotlib to requirements
-
🚀 Update UI with LFS for images and models
-
Fix: pin numpy<2.0 to stop HF runtime crash
-
🔥 Fix NumPy compatibility + processor padding
-
Cleanup: remove leftover MyInference.py to avoid import conflicts
-
🐛 Fix: switch to CPU PyTorch for HF Spaces
-
Fix: properly lowercase inference.py for Hugging Face
-
Temp rename inference file
-
🚀 Update: new UI design, classification plot, and attention improvements
-
Auto image size, working attention map
-
Fix DINOv2 processor and update requirements
-
Add id2name.json for class mapping
-
Fix num_labels=101 for Dinov2 checkpoint
-
Use facebook/dinov2-large processor instead of custom model repo
-
Add transformers deps and show classification run button
-
Add run button for classification tool
-
Fix Docker CMD and UI run button
-
Fix repo_id to Arew99/dinov2-costum for model loading
-
Fix Dockerfile to copy full app directory
-
Move Dockerfile to root for Hugging Face deployment
-
Update requirements for Hugging Face build
-
Track logo with Git LFS
-
Switch to loading model from Hugging Face Hub
-
Load model from Hugging Face Hub instead of local file
-
Remove local model weights for upload
-
Remove local model weights for upload
-
Initial commit for NEMO tools FastAPI Space
-
initial commit
Andreas Kofidis
