I had a problem due to upgraded ncsdk 2.10.
The code, shown as below, is able to run on ncsdk 2.08.
On the version of 2.10, it just run one step of loop which crash occurs at the next step.
When back to the version 2.08, the problem doesn't happend again.
How to resolve this problem in ncsdk 2.10 ?
for root, dirs, files in os.walk(path_to_images): for file in files: image_filename=root+'/'+file img = cv2.imread(image_filename).astype(np.float32) graph.queue_inference_with_fifo_elem(fifoIn, fifoOut, img, 'user object') output, userobj = fifoOut.read_elem()
The error is given by:
Number of categories: 10 /usr/local/lib/python3.5/dist-packages/mvnc/mvncapi.py:416: DeprecationWarning: The binary mode of fromstring is deprecated, as it behaves surprisingly on unicode inputs. Use frombuffer instead tensor = numpy.fromstring(tensor.raw, dtype=numpy.float32) E: [ 0] dispatcherEventReceive:247 dispatcherEventReceive() Read failed -1 E: [ 0] eventReader:268 Failed to receive event, the device may have reset E: [ 0] checkGraphMonitorResponse:922 XLink error Traceback (most recent call last): File "run.py", line 83, in <module> graph.queue_inference_with_fifo_elem(fifoIn, fifoOut, img, 'user object') File "/usr/local/lib/python3.5/dist-packages/mvnc/mvncapi.py", line 767, in queue_inference_with_fifo_elem raise Exception(Status(status)) Exception: Status.ERROR
Hi @bravo325806 ,
Thanks for reaching out. May I ask if you based your project off of the NCSDK or NCAPPZOO examples? I'd like to give this a try with the examples we provide and see if I can reproduce the problem, or if you can share your code for me to test and reproduce that would be great.
Thanks Luis answer,
I think my project not based off of the NCSDK or NCAPPZOO.
This Is my code，and I have convert the batch_size of placeholder to 1 from None：
and this my checkpoint，the checkpoint is from upper code