Replace dynamic-range int8 with weight-only int8

README.md

@@ -28,12 +28,6 @@ model-index:
           - name: Top 5 Accuracy (Full Precision)
             type: accuracy
             value: 0.9353
-          - name: Top 1 Accuracy (Dynamic Quantized wi8 afp32)
-            type: accuracy
-            value: 0.7047
-          - name: Top 5 Accuracy (Dynamic Quantized wi8 afp32)
-            type: accuracy
-            value: 0.8950
 ---
 
 # EfficientNet B0
@@ -53,6 +47,16 @@ acc@1 (on ImageNet-1K): 77.692%
 acc@5 (on ImageNet-1K): 93.532%    
 num_params: 5,288,548
 
+### Quantized variant
+
+`efficientnet_b0_weight_only_wi8_afp32.tflite` is a weight-only int8
+quantization of the same weights (about 3.5x smaller than float32).
+Weight-only quantization is used instead of dynamic-range quantization
+because EfficientNet's SE and SiLU layers are sensitive to activation
+quantization; in a spot check against the float model the weight-only
+file keeps the top-1 predictions on real photos with a minimum logit
+correlation of 0.996.
+
 ## How to Use
 
 **1. Install Dependencies** Ensure your Python environment is set up with the required libraries. Run the following command in your terminal:   

efficientnet_b0_dynamic_wi8_afp32.tflite → efficientnet_b0_weight_only_wi8_afp32.tflite

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:86551f77214b2e24779a5ed3e9532f252b5bff352ba949a90e6372739bb6b685
-size 5988912
+oid sha256:16bc1c4fa541cef382b37f204458882862468638c7ccae7f8cb6b249e4788394
+size 6032208