Update README - Run 20251012_221046

README.md

@@ -21,7 +21,7 @@ model-index:
       type: imagenet-1k
     metrics:
     - type: accuracy
-      value: 66.74
+      value: 66.84
 ---
 
 # David: Multi-Scale Feature Classifier
@@ -55,16 +55,16 @@ exist simultaneously in the same shared space with the correct checks and spacin
 ## Performance
 
 ### Best Results
-- **Validation Accuracy**: 66.74%
-- **Best Epoch**: 8
-- **Final Train Accuracy**: 65.92%
+- **Validation Accuracy**: 66.84%
+- **Best Epoch**: 9
+- **Final Train Accuracy**: 66.12%
 
 ### Per-Scale Performance
-- **Scale 256**: 66.74%
-- **Scale 512**: 72.66%
-- **Scale 768**: 74.31%
-- **Scale 1024**: 74.97%
-- **Scale 1280**: 75.24%
+- **Scale 256**: 66.84%
+- **Scale 512**: 72.72%
+- **Scale 768**: 74.34%
+- **Scale 1024**: 75.09%
+- **Scale 1280**: 75.37%
 
 
 ## Usage
@@ -85,8 +85,8 @@ AbstractPhil/david-shared-space/
 │       └── 20251012_221046/
 │           ├── MODEL_SUMMARY.txt     # 🎯 Human-readable performance summary
 │           ├── training_history.json # 📈 Epoch-by-epoch training curve
-│           ├── best_model_acc66.74.safetensors  # ⭐ Accuracy in filename!
-│           ├── best_model_acc66.74_metadata.json
+│           ├── best_model_acc66.84.safetensors  # ⭐ Accuracy in filename!
+│           ├── best_model_acc66.84_metadata.json
 │           ├── final_model.safetensors
 │           ├── checkpoint_epoch_X_accYY.YY.safetensors
 │           ├── david_config.json
@@ -108,7 +108,7 @@ from huggingface_hub import hf_hub_download
 # Specify model variant and run
 model_name = "david_high_accuracy"
 run_id = "20251012_221046"
-accuracy = "66.74"  # From MODELS_INDEX.json
+accuracy = "66.84"  # From MODELS_INDEX.json
 
 # Download config
 config_path = hf_hub_download(
@@ -217,4 +217,4 @@ Special thanks to Claude (Anthropic) for debugging assistance.
 
 ---
 
-*Generated on 2025-10-12 22:53:04*
+*Generated on 2025-10-12 22:57:47*