we have a requirement to run OpenVino sample application in OneAPI dev cloud enviroment, when we tried to run sample offline_speech_recognition_app encountering the below error.
Please do support if there is any step being missed or the account is not intended to run OpenVino application.
[ INFO ] Using feature transformation /home/u44141/openVino_speechkaldi/lspeech_s5_ext/lspeech_s5_ext.feature_transform
[ INFO ] InferenceEngine API ver. 2.1 (build: 2020.3.0-3467-15f2c61a-releases/2020/3)
[ ERROR ] Failed to initialize due to: Failed to create plugin /glob/development-tools/versions/oneapi/beta07/openvino_2020.3.194/deployment_tools/inference_engine/lib/intel64/libMKLDNNPlugin.so for device CPU
Please, check your environment
Cannot load library '/glob/development-tools/versions/oneapi/beta07/openvino_2020.3.194/deployment_tools/inference_engine/lib/intel64/libMKLDNNPlugin.so': /glob/development-tools/versions/oneapi/beta07/openvino_2020.3.194/deployment_tools/inference_engine/lib/intel64/libMKLDNNPlugin.so: failed to map segment from shared object
Failed to initialize speech library. Status: -1
Thanks for reaching out to us. Could you please share the commands used for running the openvino sample ? Also please confirm whether setupvars.sh is being sourced before running the application.
>>whether setupvars.sh is being sourced before running the application.
yes, I have sourced the below file.
>>Could you please share the commands used for running the open vino sample
I'm running intel provided open vino sample code (offline_speech_recognition_app), with the below command, have modified the code to provide wav file and config file internally instead of command line arguments.
Running in the below format
It seems like the same issue is being discussed in openvino forum. You can check further details on below link. We are closing this case from devcloud forum by @Max_L_Intel .
Can you please share some information on what OpenVINO samples have you verified in devcloud? By looking at @Daya message in this thread it seems like OpenVINO devcloud version is limited, as least from sudo user perspective.
Sudo access is usually required to install dependencies for verification scripts, and as I see in the other thread there is a workaround provided for demo_squeezenet_download_convert_run.sh
Thanks in advance.
Thank you for the support.
I'm getting below error message when tried to execute demo_speech_recognition.sh
Please find the details below:
[setupvars.sh] OpenVINO environment initialized
this server does not allow you to run "sudo".
If you need to install a dependency, please put it in your home directory.
* For autotools, you can do it with "./configure --prefix=~/my-path"
* For Python packages, use "pip install <package_name> --user" or create a Conda environment.
Error on or near line 109; exiting with status 1
Thanks for your patience. We tried running the samples in devcloud and we are getting some error which is due to lack of privileges. Also we tried editing the code by removing the sudo but no luck.
Could you please check and let us know if there is any workaround in order to run the sample in an limited access environment ?
Thanks for reaching out. The script you mentioned is trying to run commands which require sudo privileges.
If you are trying to run OpenVINO samples on Intel Devcloud for the Edge, you can follow the steps below:
1- Go to: https://devcloud.intel.com/edge/ and login using the username and password in the Welcome email you received when your Intel Devcloud for the Edge account was created.
2- If you want to run a sample tutorial, form the home page choose Tutorials.
3- If you are interested in more advanced applications like speech recognition, navigate to Advanced-> Sample Applications and choose one of the OpenVINO samples available of the Devcloud for the Edge, this will take you to a jupyter notebook with detailed explanation on how to run OpenVINO sample on one of the edge nodes.