---
language: en
license: mit
tags:
- image-classification
- imagenet
- multi-scale
- crystal-geometry
- david
datasets:
- imagenet-1k
metrics:
- accuracy
model-index:
- name: David-fully_shared-weighted_sum
  results:
  - task:
      type: image-classification
    dataset:
      name: ImageNet-1K
      type: imagenet-1k
    metrics:
    - type: accuracy
      value: 71.10
---

# David: Multi-Scale Crystal Classifier

**David** is a multi-scale deep learning classifier that uses crystal geometry (pentachora/4-simplexes) 
as class prototypes with role-weighted similarity computation (Rose Loss).

## Model Details

### Architecture
- **Preset**: small_fast
- **Sharing Mode**: fully_shared
- **Fusion Mode**: weighted_sum
- **Scales**: [256, 512]
- **Feature Dim**: 512
- **Parameters**: ~8.8M

### Training Configuration
- **Dataset**: AbstractPhil/imagenet-clip-features-orderly
- **Model Variant**: clip_vit_laion_b32
- **Epochs**: 20
- **Batch Size**: 1024
- **Learning Rate**: 0.001
- **Rose Loss Weight**: 0.1 → 0.5
- **Cayley Loss**: False

## Performance

### Best Results
- **Validation Accuracy**: 71.10%
- **Best Epoch**: 4
- **Final Train Accuracy**: 71.40%

### Per-Scale Performance
- **Scale 256**: 70.46%
- **Scale 512**: 70.96%


## Usage

### Repository Structure

```
AbstractPhil/gated-david/
├── weights/
│   └── david_small_fast/
│       └── 20251012_132646/
│           ├── best_model.safetensors
│           ├── best_model_metadata.json
│           ├── final_model.safetensors
│           ├── checkpoint_epoch_X.safetensors
│           ├── david_config.json
│           └── train_config.json
├── runs/
│   └── david_small_fast/
│       └── 20251012_132646/
│           └── events.out.tfevents.*
├── README.md
└── best_model.json
```

### Loading the Model

```python
from geovocab2.train.model.core.david import David, DavidArchitectureConfig
from huggingface_hub import hf_hub_download

# Specify model variant and run
model_name = "david_small_fast"
run_id = "20251012_132646"

# Download config
config_path = hf_hub_download(
    repo_id="AbstractPhil/gated-david", 
    filename=f"weights/{model_name}/{run_id}/david_config.json"
)
config = DavidArchitectureConfig.from_json(config_path)

# Download weights
weights_path = hf_hub_download(
    repo_id="AbstractPhil/gated-david", 
    filename=f"weights/{model_name}/{run_id}/best_model.safetensors"
)

# Load model
from safetensors.torch import load_file
david = David.from_config(config)
david.load_state_dict(load_file(weights_path))
david.eval()
```

### Inference

```python
import torch
import torch.nn.functional as F

# Assuming you have CLIP features (512-dim for ViT-B/16)
features = get_clip_features(image)  # [1, 512]

# Load anchors
anchors_dict = torch.load("anchors.pth")

# Forward pass
with torch.no_grad():
    logits, _ = david(features, anchors_dict)
    predictions = logits.argmax(dim=-1)
```

## Architecture Overview

### Multi-Scale Processing
David processes inputs at multiple scales (256, 512), 
allowing it to capture both coarse and fine-grained features.

### Crystal Geometry
Each class is represented by a pentachoron (4-simplex) in embedding space with 5 vertices:
- **Anchor**: Primary class representative
- **Need**: Complementary direction
- **Relation**: Contextual alignment
- **Purpose**: Functional direction
- **Observer**: Meta-perspective

### Rose Loss
Similarity computation uses role-weighted cosine similarities:
```
score = w_anchor * sim(z, anchor) + w_need * sim(z, need) + ...
```

### Fusion Strategy
**weighted_sum**: Intelligently combines predictions from multiple scales.

## Training Details

### Loss Components
- **Cross-Entropy**: Standard classification loss
- **Rose Loss**: Pentachora role-weighted margin loss (weight: 0.1→0.5)
- **Cayley Loss**: Geometric regularization (disabled)

### Optimization
- **Optimizer**: AdamW
- **Weight Decay**: 1e-05
- **Scheduler**: cosine_restarts
- **Gradient Clip**: 10.0
- **Mixed Precision**: False

## Citation

```bibtex
@software{david_classifier_2025,
  title = {David: Multi-Scale Crystal Classifier},
  author = {AbstractPhil},
  year = {2025},
  url = {https://huggingface.co/AbstractPhil/gated-david},
  note = {Run ID: 20251012_132646}
}
```

## License

MIT License

## Acknowledgments

Built with crystal lattice geometry and multi-scale deep learning.
Special thanks to Claude (Anthropic) for debugging assistance.

---

*Generated on 2025-10-12 13:36:37*