Fix #1, add initial net specification to avoid dealing with manual

prototxt creation.

If it reaches enough impact, I should care about the licensing of this
portion of code.

src/create_proto.py

@@ -0,0 +1,120 @@
+from collections import OrderedDict
+import re
+
+from caffe.proto import caffe_pb2
+from google import protobuf
+
+def uncamel(s):
+    """Convert CamelCase to underscore_case."""
+
+    return re.sub('(?!^)([A-Z])(?=[^A-Z])', r'_\1', s).lower()
+
+def assign_proto(proto, name, val):
+    if isinstance(val, list):
+        if isinstance(val[0], dict):
+            for item in val:
+                proto_item = getattr(proto, name).add()
+                for k, v in item.iteritems():
+                    assign_proto(proto_item, k, v)
+        else:
+            getattr(proto, name).extend(val)
+    elif isinstance(val, dict):
+        for k, v in val.iteritems():
+            assign_proto(getattr(proto, name), k, v)
+    else:
+        setattr(proto, name, val)
+
+class Top(object):
+    def __init__(self, fn, n):
+        self.fn = fn
+        self.n = n
+
+class Function(object):
+    def __init__(self, type_name, inputs, params):
+        self.type_name = type_name
+        self.inputs = inputs
+        self.params = params
+        self.ntop = self.params.get('ntop', 1)
+        if 'ntop' in self.params:
+            del self.params['ntop']
+        self.in_place = self.params.get('in_place', False)
+        if 'in_place' in self.params:
+            del self.params['in_place']
+        self.tops = tuple(Top(self, n) for n in range(self.ntop))
+
+    def _get_name(self, top, names, autonames):
+        if top not in names:
+            n = autonames.setdefault(top.fn.type_name, 1)
+            autonames[top.fn.type_name] += 1
+            names[top] = uncamel(top.fn.type_name) + str(n)
+        return names[top]
+
+    def _to_proto(self, layers, names, autonames):
+        if self in layers:
+            return
+        bottom_names = []
+        for inp in self.inputs:
+            inp.fn._to_proto(layers, names, autonames)
+            bottom_names.append(layers[inp.fn].top[inp.n])
+        layer = caffe_pb2.LayerParameter()
+        layer.type = self.type_name
+        layer.bottom.extend(bottom_names)
+
+        if self.in_place:
+            layer.top.extend(layer.bottom)
+        else:
+            for top in self.tops:
+                layer.top.append(self._get_name(top, names, autonames))
+        layer.name = self._get_name(self.tops[0], names, autonames)
+
+        for k, v in self.params.iteritems():
+            # special case to handle generic *params
+            if k.endswith('param'):
+                assign_proto(layer, k, v)
+            else:
+                try:
+                    assign_proto(getattr(layer, uncamel(self.type_name) + '_param'), k, v)
+                except AttributeError:
+                    assign_proto(layer, k, v)
+
+        layers[self] = layer
+
+class NetSpec(object):
+    def __init__(self):
+        super(NetSpec, self).__setattr__('tops', OrderedDict())
+
+    def __setattr__(self, name, value):
+        self.tops[name] = value
+
+    def __getattr__(self, name):
+        return self.tops[name]
+
+    def to_proto(self):
+        names = {v: k for k, v in self.tops.iteritems()}
+        autonames = {}
+        layers = OrderedDict()
+        for name, top in self.tops.iteritems():
+            top.fn._to_proto(layers, names, autonames)
+        net = caffe_pb2.NetParameter()
+        net.layer.extend(layers.values())
+        return net
+
+class Layers(object):
+    def __getattr__(self, name):
+        def layer_fn(*args, **kwargs):
+            fn = Function(name, args, kwargs)
+            if fn.ntop == 1:
+                return fn.tops[0]
+            else:
+                return fn.tops
+        return layer_fn
+
+class Parameters(object):
+    def __getattr__(self, name):
+       class Param:
+            def __getattr__(self, param_name):
+                return getattr(getattr(caffe_pb2, name + 'Parameter'), param_name)
+       return Param()
+
+layers = Layers()
+params = Parameters()

src/test/test_create_proto.py

@@ -0,0 +1,54 @@
+from caffe.proto import caffe_pb2
+
+from create_proto import layers as L, params as P, NetSpec
+
+# helper function for common structures
+
+def conv_relu(bottom, ks, nout, stride=1, pad=0, group=1):
+    conv = L.Convolution(bottom, kernel_size=ks, stride=stride,
+                                num_output=nout, pad=pad, group=group)
+    return conv, L.ReLU(conv, in_place=True)
+
+def fc_relu(bottom, nout):
+    fc = L.InnerProduct(bottom, num_output=nout)
+    return fc, L.ReLU(fc, in_place=True)
+
+def max_pool(bottom, ks, stride=1):
+    return L.Pooling(bottom, pool=P.Pooling.MAX, kernel_size=ks, stride=stride)
+
+def alexnet(lmdb, batch_size=256, include_acc=False):
+    net = NetSpec()
+    net.data, net.label = L.Data(source=lmdb, backend=P.Data.LMDB, batch_size=batch_size, ntop=2,
+        transform_param=dict(crop_size=227, mean_value=[104, 117, 123], mirror=True))
+
+    # the net itself
+    net.conv1, net.relu1 = conv_relu(net.data, 11, 96, stride=4)
+    net.pool1 = max_pool(net.relu1, 3, stride=2)
+    net.norm1 = L.LRN(net.pool1, local_size=5, alpha=1e-4, beta=0.75)
+    net.conv2, net.relu2 = conv_relu(net.norm1, 5, 256, pad=2, group=2)
+    net.pool2 = max_pool(net.relu2, 3, stride=2)
+    net.norm2 = L.LRN(net.pool2, local_size=5, alpha=1e-4, beta=0.75)
+    net.conv3, net.relu3 = conv_relu(net.norm2, 3, 384, pad=1)
+    net.conv4, net.relu4 = conv_relu(net.relu3, 3, 384, pad=1, group=2)
+    net.conv5, net.relu5 = conv_relu(net.relu4, 3, 256, pad=1, group=2)
+    net.pool5 = max_pool(net.relu5, 3, stride=2)
+    net.fc6, net.relu6 = fc_relu(net.pool5, 4096)
+    net.drop6 = L.Dropout(net.relu6, in_place=True)
+    net.fc7, net.relu7 = fc_relu(net.drop6, 4096)
+    net.drop7 = L.Dropout(net.relu7, in_place=True)
+    net.fc8 = L.InnerProduct(net.drop7, num_output=1000)
+    net.loss = L.SoftmaxWithLoss(net.fc8, net.label)
+
+    if include_acc:
+        net.acc = L.Accuracy(net.fc8, net.label)
+    return net.to_proto()
+
+def make_net():
+    with open('train.prototxt', 'w') as f:
+        print >>f, alexnet('/path/to/caffe-train-lmdb')
+
+    with open('test.prototxt', 'w') as f:
+        print >>f, alexnet('/path/to/caffe-val-lmdb', batch_size=50, include_acc=True)
+
+if __name__ == '__main__':
+    make_net()