diff --git a/README.md b/README.md
index f1ce6be..269ba25 100644
--- a/README.md
+++ b/README.md
@@ -13,6 +13,16 @@ The output consists of two files:
 1. A data file (in NumPy's native format) containing the model's learned parameters.
 2. A Python class that constructs the model's graph.
 
+Run `pack-pb.py` to pack code and data to pb file
+
+before packing, you need generate code and data first
+
+e.g. convert caffe alexnet to tensorflow pb file
+
+1. `./convert.py ~/caffe/models/bvlc_alexnet/deploy.prototxt --caffemodel ~/caffe/models/bvlc_alexnet/bvlc_alexnet.caffemodel --data-output-path=AlexNet.npy --code-output-path=AlexNet.py`
+2. `python pack-pb.py  --model AlexNet`
+
+
 ### Examples
 
 See the [examples](examples/) folder for more details.
diff --git a/kaffe/caffe/caffepb.py b/kaffe/caffe/caffe_pb2.py
similarity index 100%
rename from kaffe/caffe/caffepb.py
rename to kaffe/caffe/caffe_pb2.py
diff --git a/kaffe/caffe/resolver.py b/kaffe/caffe/resolver.py
index b9580a7..0986aa2 100644
--- a/kaffe/caffe/resolver.py
+++ b/kaffe/caffe/resolver.py
@@ -14,7 +14,7 @@ def import_caffe(self):
             self.caffe = caffe
         except ImportError:
             # Fall back to the protobuf implementation
-            from . import caffepb
+            from . import caffe_pb2 as caffepb
             self.caffepb = caffepb
             show_fallback_warning()
         if self.caffe:
diff --git a/kaffe/tensorflow/network.py b/kaffe/tensorflow/network.py
index 6f3b153..2750f5e 100644
--- a/kaffe/tensorflow/network.py
+++ b/kaffe/tensorflow/network.py
@@ -130,11 +130,11 @@ def conv(self,
                 output = convolve(input, kernel)
             else:
                 # Split the input into groups and then convolve each of them independently
-                input_groups = tf.split(3, group, input)
-                kernel_groups = tf.split(3, group, kernel)
+                input_groups = tf.split(axis=3, num_or_size_splits=group, value=input)
+                kernel_groups = tf.split(axis=3, num_or_size_splits=group, value=kernel)
                 output_groups = [convolve(i, k) for i, k in zip(input_groups, kernel_groups)]
                 # Concatenate the groups
-                output = tf.concat(3, output_groups)
+                output = tf.concat(axis=3, values=output_groups)
             # Add the biases
             if biased:
                 biases = self.make_var('biases', [c_o])
@@ -177,7 +177,7 @@ def lrn(self, input, radius, alpha, beta, name, bias=1.0):
 
     @layer
     def concat(self, inputs, axis, name):
-        return tf.concat(concat_dim=axis, values=inputs, name=name)
+        return tf.concat(axis=axis, values=inputs, name=name)
 
     @layer
     def add(self, inputs, name):
diff --git a/pack-pb.py b/pack-pb.py
new file mode 100644
index 0000000..bfb4fba
--- /dev/null
+++ b/pack-pb.py
@@ -0,0 +1,56 @@
+#!/usr/bin/env python
+
+import argparse
+import sys
+import tensorflow as tf
+from tensorflow.python.framework.graph_util import convert_variables_to_constants
+
+def convert(model,output,shape):
+    from alexnet import AlexNet as MyNet
+    MyNet=getattr(__import__(model),model) 
+    batch_size = 1
+    data_node = tf.placeholder(tf.float32, shape)
+    net = MyNet({'data': data_node})
+    model_dir='./'
+    with tf.Session() as sess:
+        output_graph = sess._graph
+        net.load(data_path=model+'.npy', session=sess)
+        graph = convert_variables_to_constants(sess, sess.graph_def, [output])
+        tf.train.write_graph(graph, '.', model+'.pb', as_text=False)
+
+def main():
+    input_height = 227
+    input_width = 227
+    input_channel=3
+    input_batch=1
+    model="LeNet"
+    output="prob"
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--model", help="model name")
+    parser.add_argument("--output", help="output name")
+    parser.add_argument("--input_height", type=int, help="input height")
+    parser.add_argument("--input_width", type=int, help="input width")
+    parser.add_argument("--input_channel", type=int, help="input channel")
+    parser.add_argument("--input_batch", type=int, help="input batch")
+    args = parser.parse_args()
+
+    if args.input_height:
+       input_height = args.input_height
+    if args.input_width:
+       input_width = args.input_width
+    if args.input_channel:
+       input_channel = args.input_channel
+    if args.input_batch:
+       input_batch = args.input_batch
+
+    if args.model:
+       model = args.model
+    if args.output:
+       output = args.output
+
+    convert(model,output,(input_batch,input_height,input_width,input_channel))
+
+
+if __name__ == '__main__':
+    main()