Upload runs/cifar10_weighted_20251119_023700/README.md with huggingface_hub

runs/cifar10_weighted_20251119_023700/README.md

@@ -0,0 +1,60 @@
+# Run: cifar10_weighted_20251119_023700
+
+## Configuration
+- **Dataset**: CIFAR10
+- **Fusion Mode**: weighted
+- **Parameters**: 8,936,890
+- **Simplex**: 4-simplex (5 vertices)
+
+## Performance
+- **Best Validation Accuracy**: 77.22%
+- **Training Time**: 0.7 hours
+- **Batch Size**: 128
+- **Mixed Precision**: False
+- **Final Epoch**: 100
+
+## Files
+- `runs/cifar10_weighted_20251119_023700/checkpoints/best_model.safetensors` - Model weights (SafeTensors)
+- `runs/cifar10_weighted_20251119_023700/checkpoints/best_training_state.pt` - Optimizer/scheduler state
+- `runs/cifar10_weighted_20251119_023700/checkpoints/best_metadata.json` - Training metadata
+- `runs/cifar10_weighted_20251119_023700/config.yaml` - Full configuration
+- `runs/cifar10_weighted_20251119_023700/tensorboard/` - TensorBoard logs
+
+## Usage
+```python
+from safetensors.torch import load_file
+import torch
+
+# Download from HuggingFace Hub
+from huggingface_hub import hf_hub_download
+
+model_path = hf_hub_download(
+    repo_id="AbstractPhil/vit-beans-v3",
+    filename="runs/cifar10_weighted_20251119_023700/checkpoints/best_model.safetensors"
+)
+
+# Load model weights (SafeTensors - no pickle!)
+state_dict = load_file(model_path)
+model.load_state_dict(state_dict)
+```
+
+## Training Configuration
+```yaml
+embed_dim: 384
+num_fusion_blocks: 6
+num_heads: 8
+fusion_mode: weighted
+k_simplex: 4
+learning_rate: 0.0003
+batch_size: 128
+epochs: 100
+weight_decay: 0.05
+```
+
+## Details
+
+Built with geometric consciousness-aware routing using the Devil's Staircase (Beatrix) and pentachoron parameterization.
+
+**Training completed**: 2025-11-19 03:22:18
+
+**Safe Format**: All model weights use SafeTensors (not pickle) for maximum security.