esandorfi/photo-classification
Domain features first reorganisation
Add README and STORY in Dockerfile
Home page
Change to API default directory
HF Req
Clean eval
Eval updates
Evaluation
Update Agents & Readme
Eval classify_dataset
Apply enhancements
Update README with guard usage instructions
Add scene-dance-formation-group-style-framing-guard-v3.json
Add scene-dance-formation-group-style-framing-v2 dataset
Add scene-dance-formation-group-v1 dataset
Update README with recommended runtime config
Add personal-photos-lite-v1 dataset
Add personal photos dataset JSON file
Include integration tests in pytest execution
Handle CLIP availability in integration test
Create __init__.py
Add integration test for real CLIP functionality
Add pytest configuration for integration tests
Add setup script for dependencies and testing
Add development dependencies for pytest and httpx
Implement FakeClipStore and FakeTwoStageClassifier
Integrate fake store and classifier in test client
Add tests for label set API endpoints
Add pytest fixtures for testing FastAPI app
Refactor app initialization into build_app function
Implement create_app function in app_factory.py
Enhance classify function with request and logging
Add Photo Classification API with label set endpoints
Add dependency functions for label set management
Implement ClipScorer and TwoStageClassifier classes
Add StageTimings and ClassificationResult data classes
Implement ClipStore class for CLIP model handling
Implement LabelSetRegistry with upsert and activate methods
Add EmbeddingBank and LabelSetBank data classes
Add function to load image from base64 string
Implement RequestIdMiddleware for request tracking
Add logging utility functions for JSON logging
Add stable_hash function for JSON objects
Create Pydantic models for label management
Add settings configuration using Pydantic
Add dependencies for FastAPI and machine learning
Update README.md with project structure
Change base image from Python 3.13 to 3.12
Add project structure to README.md