Update README - Run 20251104_122236

README.md

@@ -5,14 +5,14 @@ tags:
 - image-classification
 - imagenet
 - multi-scale
-- crystal-geometry
+- feature-geometry
 - david
 datasets:
 - imagenet-1k
 metrics:
 - accuracy
 model-index:
-- name: David-partial_shared-deep_efficiency
+- name: David-partial_shared-geometric_attention
   results:
   - task:
       type: image-classification
@@ -21,28 +21,28 @@ model-index:
       type: imagenet-1k
     metrics:
     - type: accuracy
-      value: 84.73
+      value: 76.11
 ---
 
-# David: Multi-Scale Crystal Classifier
+# David: Multi-Scale Feature Classifier
 
-**David** is a multi-scale deep learning classifier that uses crystal geometry (pentachora/4-simplexes) 
+**David** is a multi-scale deep learning classifier that uses feature geometry (pentachora/4-simplexes)
 as class prototypes with role-weighted similarity computation (Rose Loss).
 
 ## Model Details
 
 ### Architecture
-- **Preset**: clip_vit_bigg14
+- **Preset**: clip_vit_b16_geometric
 - **Sharing Mode**: partial_shared
-- **Fusion Mode**: deep_efficiency
-- **Scales**: [384, 512, 768, 1024, 1280, 1536, 1792, 2048]
-- **Feature Dim**: 1280
-- **Parameters**: 60,814,373
+- **Fusion Mode**: geometric_attention
+- **Scales**: [256, 512, 768, 1024]
+- **Feature Dim**: 512
+- **Parameters**: 12,599,556
 
 ### Training Configuration
 - **Dataset**: AbstractPhil/imagenet-clip-features-orderly
-- **Model Variant**: clip_vit_laion_bigg14
-- **Epochs**: 10
+- **Model Variant**: clip_vit_b16
+- **Epochs**: 5
 - **Batch Size**: 1024
 - **Learning Rate**: 0.001
 - **Rose Loss Weight**: 0.1 → 0.5
@@ -51,19 +51,15 @@ as class prototypes with role-weighted similarity computation (Rose Loss).
 ## Performance
 
 ### Best Results
-- **Validation Accuracy**: 84.73%
-- **Best Epoch**: 2
-- **Final Train Accuracy**: 96.34%
+- **Validation Accuracy**: 76.11%
+- **Best Epoch**: 0
+- **Final Train Accuracy**: 71.75%
 
 ### Per-Scale Performance
-- **Scale 384**: 84.39%
-- **Scale 512**: 84.67%
-- **Scale 768**: 84.52%
-- **Scale 1024**: 84.71%
-- **Scale 1280**: 84.73%
-- **Scale 1536**: 84.57%
-- **Scale 1792**: 84.58%
-- **Scale 2048**: 84.59%
+- **Scale 256**: 74.25%
+- **Scale 512**: 75.16%
+- **Scale 768**: 75.26%
+- **Scale 1024**: 75.82%
 
 
 ## Usage
@@ -80,43 +76,24 @@ AbstractPhil/gated-david/
 ├── README.md                         # This file
 ├── best_model.json                   # Latest best model info
 ├── weights/
-│   └── david_clip_vit_bigg14/
-│       └── 20251012_181608/
+│   └── david_clip_vit_b16_geometric/
+│       └── 20251104_122236/
 │           ├── MODEL_SUMMARY.txt     # 🎯 Human-readable performance summary
 │           ├── training_history.json # 📈 Epoch-by-epoch training curve
-│           ├── best_model_acc84.73.safetensors  # ⭐ Accuracy in filename!
-│           ├── best_model_acc84.73_metadata.json
+│           ├── best_model_acc76.11.safetensors  # ⭐ Accuracy in filename!
+│           ├── best_model_acc76.11_metadata.json
 │           ├── final_model.safetensors
 │           ├── checkpoint_epoch_X_accYY.YY.safetensors
 │           ├── david_config.json
 │           └── train_config.json
 └── runs/
-    └── david_clip_vit_bigg14/
-        └── 20251012_181608/
+    └── david_clip_vit_b16_geometric/
+        └── 20251104_122236/
             └── events.out.tfevents.* # TensorBoard logs
 ```
 
 ### Loading the Model
 
-The description below is terrible.
-
-Load a clip-vit and then load one of the pretrains. There is no anchors file nor any of that anymore.
-
-I don't know why this piece of the readme got preserved, but I'll need to fix it.
-
-1. Install the repo in your environment
-
-```python
-try:
-  !pip uninstall -qy geometricvocab
-except:
-  pass
-
-!pip install -q git+https://github.com/AbstractEyes/lattice_vocabulary.git
-```
-
-Find your target model, likely not all of this below will work currently.
-
 ```python
 from geovocab2.train.model.core.david import David, DavidArchitectureConfig
 from huggingface_hub import hf_hub_download
@@ -124,26 +101,26 @@ from huggingface_hub import hf_hub_download
 # Browse available models in MODELS_INDEX.json first!
 
 # Specify model variant and run
-model_name = "david_clip_vit_bigg14"
-run_id = "20251012_181608"
-accuracy = "84.73"  # From MODELS_INDEX.json
+model_name = "david_clip_vit_b16_geometric"
+run_id = "20251104_122236"
+accuracy = "76.11"  # From MODELS_INDEX.json
 
 # Download config
 config_path = hf_hub_download(
-    repo_id="AbstractPhil/gated-david", 
+    repo_id="AbstractPhil/gated-david",
     filename=f"weights/{model_name}/{run_id}/david_config.json"
 )
 config = DavidArchitectureConfig.from_json(config_path)
 
 # Download weights (accuracy in filename!)
 weights_path = hf_hub_download(
-    repo_id="AbstractPhil/gated-david", 
+    repo_id="AbstractPhil/gated-david",
     filename=f"weights/{model_name}/{run_id}/best_model_acc{accuracy}.safetensors"
 )
 
 # Download training history (optional - see full training curve)
 history_path = hf_hub_download(
-    repo_id="AbstractPhil/gated-david", 
+    repo_id="AbstractPhil/gated-david",
     filename=f"weights/{model_name}/{run_id}/training_history.json"
 )
 
@@ -154,18 +131,31 @@ david.load_state_dict(load_file(weights_path))
 david.eval()
 ```
 
-3. When in eval mode pass the output from clip-vit's feature into david. The feature does not require the token nor the label, simply pass the feature in and assess the outputs.
+### 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]
 
-  I'll need to ensure this process is perfected later, so the trainings can actually be utilized.
+# 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 (384, 512, 768, 1024, 1280, 1536, 1792, 2048), 
+David processes inputs at multiple scales (256, 512, 768, 1024),
 allowing it to capture both coarse and fine-grained features.
 
-### Crystal Geometry
+### Feature Geometry
 Each class is represented by a pentachoron (4-simplex) in embedding space with 5 vertices:
 - **Anchor**: Primary class representative
 - **Need**: Complementary direction
@@ -180,7 +170,7 @@ score = w_anchor * sim(z, anchor) + w_need * sim(z, need) + ...
 ```
 
 ### Fusion Strategy
-**deep_efficiency**: Intelligently combines predictions from multiple scales.
+**geometric_attention**: Intelligently combines predictions from multiple scales.
 
 ## Training Details
 
@@ -200,11 +190,11 @@ score = w_anchor * sim(z, anchor) + w_need * sim(z, need) + ...
 
 ```bibtex
 @software{david_classifier_2025,
-  title = {David: Multi-Scale Crystal Classifier},
+  title = {David: Multi-Scale Feature Classifier},
   author = {AbstractPhil},
   year = {2025},
   url = {https://huggingface.co/AbstractPhil/gated-david},
-  note = {Run ID: 20251012_181608}
+  note = {Run ID: 20251104_122236}
 }
 ```
 
@@ -214,9 +204,9 @@ MIT License
 
 ## Acknowledgments
 
-Built with crystal lattice geometry and multi-scale deep learning.
+Built with lattice geometry and multi-scale deep learning.
 Special thanks to Claude (Anthropic) for debugging assistance.
 
 ---
 
-*Generated on 2025-10-12 18:49:29*
+*Generated on 2025-11-04 12:24:29*