AbstractPhil commited on
Commit
7cda0a5
Β·
verified Β·
1 Parent(s): 4538b2c

Update README - Run 20251012_221046

Browse files
Files changed (1) hide show
  1. README.md +13 -13
README.md CHANGED
@@ -21,7 +21,7 @@ model-index:
21
  type: imagenet-1k
22
  metrics:
23
  - type: accuracy
24
- value: 66.14
25
  ---
26
 
27
  # David: Multi-Scale Feature Classifier
@@ -55,16 +55,16 @@ exist simultaneously in the same shared space with the correct checks and spacin
55
  ## Performance
56
 
57
  ### Best Results
58
- - **Validation Accuracy**: 66.14%
59
- - **Best Epoch**: 5
60
- - **Final Train Accuracy**: 64.71%
61
 
62
  ### Per-Scale Performance
63
- - **Scale 256**: 66.14%
64
- - **Scale 512**: 72.07%
65
- - **Scale 768**: 73.59%
66
- - **Scale 1024**: 73.95%
67
- - **Scale 1280**: 73.67%
68
 
69
 
70
  ## Usage
@@ -85,8 +85,8 @@ AbstractPhil/david-shared-space/
85
  β”‚ └── 20251012_221046/
86
  β”‚ β”œβ”€β”€ MODEL_SUMMARY.txt # 🎯 Human-readable performance summary
87
  β”‚ β”œβ”€β”€ training_history.json # πŸ“ˆ Epoch-by-epoch training curve
88
- β”‚ β”œβ”€β”€ best_model_acc66.14.safetensors # ⭐ Accuracy in filename!
89
- β”‚ β”œβ”€β”€ best_model_acc66.14_metadata.json
90
  β”‚ β”œβ”€β”€ final_model.safetensors
91
  β”‚ β”œβ”€β”€ checkpoint_epoch_X_accYY.YY.safetensors
92
  β”‚ β”œβ”€β”€ david_config.json
@@ -108,7 +108,7 @@ from huggingface_hub import hf_hub_download
108
  # Specify model variant and run
109
  model_name = "david_high_accuracy"
110
  run_id = "20251012_221046"
111
- accuracy = "66.14" # From MODELS_INDEX.json
112
 
113
  # Download config
114
  config_path = hf_hub_download(
@@ -217,4 +217,4 @@ Special thanks to Claude (Anthropic) for debugging assistance.
217
 
218
  ---
219
 
220
- *Generated on 2025-10-12 22:38:53*
 
21
  type: imagenet-1k
22
  metrics:
23
  - type: accuracy
24
+ value: 66.39
25
  ---
26
 
27
  # David: Multi-Scale Feature Classifier
 
55
  ## Performance
56
 
57
  ### Best Results
58
+ - **Validation Accuracy**: 66.39%
59
+ - **Best Epoch**: 6
60
+ - **Final Train Accuracy**: 65.19%
61
 
62
  ### Per-Scale Performance
63
+ - **Scale 256**: 66.39%
64
+ - **Scale 512**: 72.24%
65
+ - **Scale 768**: 73.89%
66
+ - **Scale 1024**: 74.43%
67
+ - **Scale 1280**: 74.45%
68
 
69
 
70
  ## Usage
 
85
  β”‚ └── 20251012_221046/
86
  β”‚ β”œβ”€β”€ MODEL_SUMMARY.txt # 🎯 Human-readable performance summary
87
  β”‚ β”œβ”€β”€ training_history.json # πŸ“ˆ Epoch-by-epoch training curve
88
+ β”‚ β”œβ”€β”€ best_model_acc66.39.safetensors # ⭐ Accuracy in filename!
89
+ β”‚ β”œβ”€β”€ best_model_acc66.39_metadata.json
90
  β”‚ β”œβ”€β”€ final_model.safetensors
91
  β”‚ β”œβ”€β”€ checkpoint_epoch_X_accYY.YY.safetensors
92
  β”‚ β”œβ”€β”€ david_config.json
 
108
  # Specify model variant and run
109
  model_name = "david_high_accuracy"
110
  run_id = "20251012_221046"
111
+ accuracy = "66.39" # From MODELS_INDEX.json
112
 
113
  # Download config
114
  config_path = hf_hub_download(
 
217
 
218
  ---
219
 
220
+ *Generated on 2025-10-12 22:43:36*