URL: https://docs.radxa.com/cubie/a7z/app-dev/npu-dev/model-zoo/yolov8
I tried running the inference as mentioned in the guide, but in the end I am getting 283 detections on the dog.jpg image only. Does anyone know what could be the reason?
Board: Radxa Cubie A7Z — Allwinner A733, 1GB RAM, 64GB microSD
OS: radxa-a733_bullseye_kde_r2 (Debian 11, kernel 5.15.147-12-a733)
Output:
radxa@radxa-cubie-a7z:~/board_deploy/yolov8_demo_linux_a733$ ./yolov8_demo_a733 -nb model/yolov8n_6_uint8_a733.nb -i model/dog.jpg
model_file=model/yolov8n_6_uint8_a733.nb, input=model/dog.jpg, loop_count=1, malloc_mbyte=10
VIPLite driver software version 2.0.3.2-AW-2024-08-30
input 0 dim 3 640 640 1, data_format=2, quant_format=0, name=input/output[0], none-quant
output 0 dim 80 80 64 1, data_format=0, name=uid_20000_sub_uid_1_out_0, none-quant
output 1 dim 80 80 64 1, data_format=0, name=uid_20001_sub_uid_1_out_0, none-quant
output 2 dim 40 40 64 1, data_format=0, name=uid_20002_sub_uid_1_out_0, none-quant
output 3 dim 40 40 64 1, data_format=0, name=uid_20003_sub_uid_1_out_0, none-quant
output 4 dim 20 20 64 1, data_format=0, name=uid_20004_sub_uid_1_out_0, none-quant
output 5 dim 20 20 64 1, data_format=0, name=uid_20005_sub_uid_1_out_0, none-quant
nbg name=model/yolov8n_6_uint8_a733.nb, size: 2127096.
create network 0: 13208 us.
prepare network: 2230 us.
buffer ptr: 0x1395e200, buffer size: 1228800
network: 0, loop count: 1
run time for this network 0: 11961 us.
detection num: 283
4: 100%, [ 206, 0, 767, 575], airplane
11: 100%, [ 0, 72, 678, 524], stop_sign
45: 100%, [ 306, 24, 767, 560], bowl
54: 100%, [ 0, 154, 558, 575], donut
43: 100%, [ 223, 201, 767, 575], knife
....
....
....
53: 100%, [ 338, 575, 598, 575], pizza
65: 100%, [ 685, 575, 767, 575], remote
67: 100%, [ 626, 575, 767, 575], cell_phone
53: 100%, [ 219, 575, 483, 575], pizza
53: 100%, [ 108, 575, 371, 575], pizza
destory npu finished.
~NpuUint.
Note: I build it on my Linux machine via WSL
Can anyone guide me on this?
Time: 4/20/2026, 3:04:29 PM
URL: https://docs.radxa.com/cubie/a7z/app-dev/npu-dev/model-zoo/yolov8
I tried running the inference as mentioned in the guide, but in the end I am getting 283 detections on the dog.jpg image only. Does anyone know what could be the reason?
Board: Radxa Cubie A7Z — Allwinner A733, 1GB RAM, 64GB microSD
OS: radxa-a733_bullseye_kde_r2 (Debian 11, kernel 5.15.147-12-a733)
Output:
radxa@radxa-cubie-a7z:~/board_deploy/yolov8_demo_linux_a733$ ./yolov8_demo_a733 -nb model/yolov8n_6_uint8_a733.nb -i model/dog.jpg
model_file=model/yolov8n_6_uint8_a733.nb, input=model/dog.jpg, loop_count=1, malloc_mbyte=10
VIPLite driver software version 2.0.3.2-AW-2024-08-30
input 0 dim 3 640 640 1, data_format=2, quant_format=0, name=input/output[0], none-quant
output 0 dim 80 80 64 1, data_format=0, name=uid_20000_sub_uid_1_out_0, none-quant
output 1 dim 80 80 64 1, data_format=0, name=uid_20001_sub_uid_1_out_0, none-quant
output 2 dim 40 40 64 1, data_format=0, name=uid_20002_sub_uid_1_out_0, none-quant
output 3 dim 40 40 64 1, data_format=0, name=uid_20003_sub_uid_1_out_0, none-quant
output 4 dim 20 20 64 1, data_format=0, name=uid_20004_sub_uid_1_out_0, none-quant
output 5 dim 20 20 64 1, data_format=0, name=uid_20005_sub_uid_1_out_0, none-quant
nbg name=model/yolov8n_6_uint8_a733.nb, size: 2127096.
create network 0: 13208 us.
prepare network: 2230 us.
buffer ptr: 0x1395e200, buffer size: 1228800
network: 0, loop count: 1
run time for this network 0: 11961 us.
detection num: 283
4: 100%, [ 206, 0, 767, 575], airplane
11: 100%, [ 0, 72, 678, 524], stop_sign
45: 100%, [ 306, 24, 767, 560], bowl
54: 100%, [ 0, 154, 558, 575], donut
43: 100%, [ 223, 201, 767, 575], knife
....
....
....
53: 100%, [ 338, 575, 598, 575], pizza
65: 100%, [ 685, 575, 767, 575], remote
67: 100%, [ 626, 575, 767, 575], cell_phone
53: 100%, [ 219, 575, 483, 575], pizza
53: 100%, [ 108, 575, 371, 575], pizza
destory npu finished.
~NpuUint.
Note: I build it on my Linux machine via WSL
Can anyone guide me on this?
Time: 4/20/2026, 3:04:29 PM