New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
NNAPI on Kirin 980 NPU. Model won't run on NPU, libc : Access denied finding property "ro.hardware.chipname" #52914
Comments
Update: Appears to be a problem of benchmark_model not having the correct permissions because when i have executed the same model inside a custom app, i don't have the problem. |
@miaowang14 Can you take a look? |
Still get the same error when using tensorfow and kotlin |
@alexymumo , could you share more details about your model and the error logs? For floating point models, you would have to set allow_fp16 to true to run it on the NPU: https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/delegates/nnapi/nnapi_delegate.h#L97 |
This issue is stale because it has been open for 7 days with no activity. It will be closed if no further activity occurs. Thank you. |
This issue was closed because it has been inactive for 7 days since being marked as stale. Please reopen if you'd like to work on this further. |
System information
I have been experiencing NNAPI producing longer inference times on my device than with standard CPU delegate, so i ran benchmark binary with tensorflow prebuilt NNAPI models. The debug from logcat shows that it get stuck with the following error
I have tested this on Huawei P30 Pro and Huawei Nova 5T with the same results.
I downloaded mobilenet_v1_1.0_224_quant.tflite NNAPI prebuilt model from here https://www.tensorflow.org/lite/performance/nnapi#use_supported_models_and_ops
Then i pushed to the device and invoked the benchmarking tool
.downloaded android aarch64 benchmark tool as per instructions and invoked the model like so
I enabled verbose nnapi log using this command
adb shell setprop debug.nn.vlog 1
To confirm that NPU exists and is available
The text was updated successfully, but these errors were encountered: