frame

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Sign In

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Status Error during Inference

Hi to everyone!!

I made a network with a very simple structure:

input = Input(shape=(1, 28, 3), name='input')
x = Conv2D(filters=1, kernel_size=(1,1), use_bias=False, strides=(1,1), padding='valid', name='conv2d')(input)
x = Flatten()(x)
x = initial_model(x)

where intial_model is a feed forward fully connected network with input 28 (coming from the previous part) and 5 as output.

I was able to export the .graph file, but when I make inference I get the following error:

" E: [ 0] dispatcherEventReceive:246 dispatcherEventReceive() Read failed -2

E: [ 0] eventReader:267 Failed to receive event, the device may have reset

Traceback (most recent call last):
File "main.py", line 143, in
app.start()
File "main.py", line 122, in start
output, user_obj = output_fifo.read_elem()
File "/usr/local/lib/python3.5/dist-packages/mvnc/mvncapi.py", line 412, in read_elem
raise Exception(Status(status))
Exception: Status.ERROR"

To debug any type of error I'm simply simulating a possible input with the following code:

input_final = np.ones((1,1,28,3)).astype(np.float32)
graph.queue_inference_with_fifo_elem(input_fifo, output_fifo, input_final,  'obj user')
output, user_obj = output_fifo.read_elem()

Do you have any idea what's going on?

Thank you in advance for your help!

N.B: if I leave the code running without output_fifo.read_elem() method, it runs for 3/4 seconds and then report a similar error, but on line 767.

Comments

This discussion has been closed.