Bi-Real Net: fix layer bug and point to updated pretrained weights (#72)

* fix layer bug and point to updated pretrained weights * fix netron link
larq · Sep 8, 2019 · f6bf31c · f6bf31c
1 parent 316f387
commit f6bf31c
Showing 1 changed file with 11 additions and 9 deletions.
diff --git a/larq_zoo/birealnet.py b/larq_zoo/birealnet.py
@@ -52,15 +52,17 @@ def residual_block(x, double_filters=False, filters=None):
     out = tf.keras.layers.BatchNormalization(momentum=0.8)(out)
     out = tf.keras.layers.MaxPool2D(3, strides=2, padding="same")(out)
 
-    # layer 2 - 5
+    # layer 2
     out = residual_block(out, filters=args.filters)
-    for _ in range(1, 5):
+
+    # layer 3 - 5
+    for _ in range(3):
         out = residual_block(out)
 
     # layer 6 - 17
-    for i in range(1, 4):
+    for _ in range(3):
         out = residual_block(out, double_filters=True)
-        for _ in range(1, 4):
+        for _ in range(3):
             out = residual_block(out)
 
     # layer 18
@@ -108,7 +110,7 @@ def BiRealNet(
     Optionally loads weights pre-trained on ImageNet.
 
     ```netron
-    birealnet-v0.2.0/birealnet.json
+    birealnet-v0.3.0/birealnet.json
     ```
     ```plot-altair
     /plots/birealnet.vg.json
@@ -153,16 +155,16 @@ def BiRealNet(
         if include_top:
             weights_path = utils.download_pretrained_model(
                 model="birealnet",
-                version="v0.2.0",
+                version="v0.3.0",
                 file="birealnet_weights.h5",
-                file_hash="e8b29d6204663997dded5629804c0c2e309ec422512a54a17d98802fb39415ec",
+                file_hash="6e6efac1584fcd60dd024198c87f42eb53b5ec719a5ca1f527e1fe7e8b997117",
             )
         else:
             weights_path = utils.download_pretrained_model(
                 model="birealnet",
-                version="v0.2.0",
+                version="v0.3.0",
                 file="birealnet_weights_notop.h5",
-                file_hash="746ff2d2d2b794226e66f0fa3fd0ff19db836df5a9ea9a0f7e59a724e1364757",
+                file_hash="5148b61c0c2a1094bdef811f68bf4957d5ba5f83ad26437b7a4a6855441ab46b",
             )
         model.load_weights(weights_path)
     elif weights is not None: