serjl

I found a way to do it through python. Import your frozen *.pb in python and run over all nodes and rename them all to the desired GPU. Save it...

Later today will put here. But just to stress, it does work one session *.pb per device. So if you want to run on 2 gpus you need to initiate...

@AldoZ Here the Python code for two GPU's: detection_graph = tf.Graph() with detection_graph.as_default(): od_graph_def = tf.GraphDef() with tf.gfile.GFile(PATH_TO_PB, 'rb') as fid: serialized_graph = fid.read() od_graph_def.ParseFromString(serialized_graph) for i in range(2): for...

@hillin Cool, thank you very much! Regarding the return in CreateSession: shouln't it be `return new TFSession(graph, options); ` otherwise I don't see where you use the graph variable.