Update README - Run 20251012_135249

README.md

@@ -21,7 +21,7 @@ model-index:
       type: imagenet-1k
     metrics:
     - type: accuracy
-      value: 71.10
+      value: 68.06
 ---
 
 # David: Multi-Scale Crystal Classifier
@@ -37,12 +37,12 @@ as class prototypes with role-weighted similarity computation (Rose Loss).
 - **Fusion Mode**: weighted_sum
 - **Scales**: [256, 512]
 - **Feature Dim**: 512
-- **Parameters**: ~8.8M
+- **Parameters**: 656,898
 
 ### Training Configuration
 - **Dataset**: AbstractPhil/imagenet-clip-features-orderly
 - **Model Variant**: clip_vit_laion_b32
-- **Epochs**: 20
+- **Epochs**: 10
 - **Batch Size**: 1024
 - **Learning Rate**: 0.001
 - **Rose Loss Weight**: 0.1 → 0.5
@@ -51,36 +51,43 @@ as class prototypes with role-weighted similarity computation (Rose Loss).
 ## Performance
 
 ### Best Results
-- **Validation Accuracy**: 71.10%
-- **Best Epoch**: 4
-- **Final Train Accuracy**: 71.40%
+- **Validation Accuracy**: 68.06%
+- **Best Epoch**: 0
+- **Final Train Accuracy**: 63.40%
 
 ### Per-Scale Performance
-- **Scale 256**: 70.46%
-- **Scale 512**: 70.96%
+- **Scale 256**: 66.92%
+- **Scale 512**: 67.88%
 
 
 ## Usage
 
+### Quick Model Lookup
+
+**Check `MODELS_INDEX.json` in the repo root** - it lists all trained models sorted by accuracy with links to weights and configs.
+
 ### Repository Structure
 
 ```
 AbstractPhil/gated-david/
+├── MODELS_INDEX.json                # 📊 Master index of all models (sorted by accuracy)
+├── README.md                         # This file
+├── best_model.json                   # Latest best model info
 ├── weights/
 │   └── david_small_fast/
-│       └── 20251012_132646/
-│           ├── best_model.safetensors
-│           ├── best_model_metadata.json
+│       └── 20251012_135249/
+│           ├── MODEL_SUMMARY.txt     # 🎯 Human-readable performance summary
+│           ├── training_history.json # 📈 Epoch-by-epoch training curve
+│           ├── best_model_acc68.06.safetensors  # ⭐ Accuracy in filename!
+│           ├── best_model_acc68.06_metadata.json
 │           ├── final_model.safetensors
-│           ├── checkpoint_epoch_X.safetensors
+│           ├── checkpoint_epoch_X_accYY.YY.safetensors
 │           ├── david_config.json
 │           └── train_config.json
-├── runs/
-│   └── david_small_fast/
-│       └── 20251012_132646/
-│           └── events.out.tfevents.*
-├── README.md
-└── best_model.json
+└── runs/
+    └── david_small_fast/
+        └── 20251012_135249/
+            └── events.out.tfevents.* # TensorBoard logs
 ```
 
 ### Loading the Model
@@ -89,9 +96,12 @@ AbstractPhil/gated-david/
 from geovocab2.train.model.core.david import David, DavidArchitectureConfig
 from huggingface_hub import hf_hub_download
 
+# Browse available models in MODELS_INDEX.json first!
+
 # Specify model variant and run
 model_name = "david_small_fast"
-run_id = "20251012_132646"
+run_id = "20251012_135249"
+accuracy = "68.06"  # From MODELS_INDEX.json
 
 # Download config
 config_path = hf_hub_download(
@@ -100,10 +110,16 @@ config_path = hf_hub_download(
 )
 config = DavidArchitectureConfig.from_json(config_path)
 
-# Download weights
+# Download weights (accuracy in filename!)
 weights_path = hf_hub_download(
     repo_id="AbstractPhil/gated-david", 
-    filename=f"weights/{model_name}/{run_id}/best_model.safetensors"
+    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", 
+    filename=f"weights/{model_name}/{run_id}/training_history.json"
 )
 
 # Load model
@@ -176,7 +192,7 @@ score = w_anchor * sim(z, anchor) + w_need * sim(z, need) + ...
   author = {AbstractPhil},
   year = {2025},
   url = {https://huggingface.co/AbstractPhil/gated-david},
-  note = {Run ID: 20251012_132646}
+  note = {Run ID: 20251012_135249}
 }
 ```
 
@@ -191,4 +207,4 @@ Special thanks to Claude (Anthropic) for debugging assistance.
 
 ---
 
-*Generated on 2025-10-12 13:36:37*
+*Generated on 2025-10-12 13:54:34*