Community
cancel
Showing results for 
Search instead for 
Did you mean: 
idata
Community Manager
375 Views

Model that does nothing

I am trying to build a model for the NCS that does nothing. When compiling the model I get an error

 

shape: (400, 400) res.shape: (400, 400) TensorFlow output shape: (1, 400, 400) Traceback (most recent call last): File "/usr/local/bin/mvNCCompile", line 169, in <module> create_graph(args.network, args.image, args.inputnode, args.outputnode, args.outfile, args.nshaves, args.inputsize, args.weights, args.explicit_concat, args.ma2480, args.scheduler, args.new_parser, args) File "/usr/local/bin/mvNCCompile", line 148, in create_graph load_ret = load_network(args, parser, myriad_config) File "/usr/local/bin/ncsdk/Controllers/Scheduler.py", line 110, in load_network network.optimize() File "/usr/local/bin/ncsdk/Models/Network.py", line 250, in optimize self.convert_network_input_to_yxz() File "/usr/local/bin/ncsdk/Models/Network.py", line 337, in convert_network_input_to_yxz if self.stageslist[0].op in [StageType.fully_connected_layer, StageType.convolution, StageType.max_pooling, IndexError: list index out of range

 

My model is build as follows:

 

import tensorflow as tf import numpy as np def main(_): graph = tf.Graph() with graph.as_default(): with tf.Session(graph=graph) as sess: x = tf.placeholder(tf.float32, [400, 400], name='input') for n in tf.get_default_graph().as_graph_def().node: print(n.name + ": " + n.op) sess.run(tf.global_variables_initializer()) sess.run(tf.local_variables_initializer()) # test model in_tensor = tf.get_default_graph().get_tensor_by_name('input:0') input_data = np.random.rand(400, 400).astype(np.float32) known = sess.run('input:0', {in_tensor: input_data}) print('Result: ' + str(known)) # attempts to use saver but fails due to no variables to save #saver = tf.train.Saver(tf.global_variables()) #saver.save(sess, './empty') tf.train.write_graph(sess.graph_def, './', 'empty.pb', as_text=False) tf.app.run(main=main)

 

Any idea on how to get past this hurdle?

Tags (1)
0 Kudos
2 Replies
idata
Community Manager
76 Views

oh bro,

 

you can use mvNCCheck tools to check if your model has been translate successfully,

 

the tool's documentation in this site : https://movidius.github.io/ncsdk/tools/check.html
idata
Community Manager
76 Views

bro, mvNCCheck gives me the same error…

 

my guess is it doesn't like that the model doesn't have any weights

Reply