.

optim_nn.py

@@ -65,6 +65,12 @@ class LayerNorm(Layer):
         self.affine = bool(affine)
         self.size = None
 
+        if self.affine:
+            self.serialized = {
+                'gamma': 'gamma',
+                'beta': 'beta',
+            }
+
     def make_shape(self, shape):
         super().make_shape(shape)
         if len(shape) != 1:
@@ -116,6 +122,11 @@ class LayerNorm(Layer):
 class Denses(Layer): # TODO: rename?
     # acts as a separate Dense for each row or column. only for 2D arrays.
 
+    serialized = {
+        'W': 'coeffs',
+        'b': 'biases',
+    }
+
     def __init__(self, dim, init=init_he_uniform, axis=-1):
         super().__init__()
         self.dim = int(dim)

optim_nn_core.py

@@ -515,6 +515,11 @@ class Softmax(Layer):
 # Parametric Layers {{{1
 
 class Dense(Layer):
+    serialized = {
+        'W': 'coeffs',
+        'b': 'biases',
+    }
+
     def __init__(self, dim, init=init_he_uniform):
         super().__init__()
         self.dim = int(dim)
@@ -615,10 +620,9 @@ class Model:
 
     def load_weights(self, fn):
         # seemingly compatible with keras' Dense layers.
-        # ignores any non-Dense layer types.
-        # TODO: assert file actually exists
         import h5py
-        f = h5py.File(fn)
+        open(fn) # just ensure the file exists (python's error is better)
+        f = h5py.File(fn, 'r')
         weights = {}
         def visitor(name, obj):
             if isinstance(obj, h5py.Dataset):
@@ -626,28 +630,42 @@ class Model:
         f.visititems(visitor)
         f.close()
 
-        denses = [node for node in self.ordered_nodes if isinstance(node, Dense)]
+        used = {}
-        for i in range(len(denses)):
+        for k in weights.keys():
-            a, b = i, i + 1
+            used[k] = False
-            b_name = "dense_{}".format(b)
+
-            # TODO: write a Dense method instead of assigning directly
+        nodes = [node for node in self.ordered_nodes if node.size is not None]
-            denses[a].coeffs[:] = weights[b_name+'_W']
+        for node in nodes:
-            denses[a].biases[:] = np.expand_dims(weights[b_name+'_b'], 0)
+            full_name = str(node).lower()
+            for s_name, o_name in node.serialized.items():
+                key = full_name + '_' + s_name
+                data = weights[key]
+                target = getattr(node, o_name)
+                target[:] = data
+                used[key] = True
+
+        for k, v in used.items():
+            if not v:
+                lament("WARNING: unused weight", k)
 
     def save_weights(self, fn, overwrite=False):
         import h5py
         f = h5py.File(fn, 'w')
 
-        denses = [node for node in self.ordered_nodes if isinstance(node, Dense)]
+        counts = defaultdict(lambda: 0)
-        for i in range(len(denses)):
+
-            a, b = i, i + 1
+        nodes = [node for node in self.ordered_nodes if node.size is not None]
-            b_name = "dense_{}".format(b)
+        for node in nodes:
-            # TODO: write a Dense method instead of assigning directly
+            full_name = str(node).lower()
-            grp = f.create_group(b_name)
+            grp = f.create_group(full_name)
-            data = grp.create_dataset(b_name+'_W', denses[a].coeffs.shape, dtype=_f)
+            for s_name, o_name in node.serialized.items():
-            data[:] = denses[a].coeffs
+                key = full_name + '_' + s_name
-            data = grp.create_dataset(b_name+'_b', denses[a].biases.shape, dtype=_f)
+                target = getattr(node, o_name)
-            data[:] = denses[a].biases
+                data = grp.create_dataset(key, target.shape, dtype=_f)
+                data[:] = target
+                counts[key] += 1
+                if counts[key] > 1:
+                    lament("WARNING: rewrote weight", key)
 
         f.close()