Created TensorFlow Lite XNNPACK delegate for CPU

Robot Framework
1

Hello peeps,

Hope everything is going well.

I am an absolute beginner in robot framework. I started writing the code for entering username and password for Browser Authentication Pop-up. Although I try using this kind of URL “https://username:password@…”, the browser is not opening due to the security reason.
I tried alternative ways like Autoit, pyautogui, javascript code. But getting this error message for any kind of option.

Created TensorFlow Lite XNNPACK delegate for CPU.
Attempting to use a delegate that only supports static-sized tensors with a graph that has dynamic-sized tensors (tensor#58 is a dynamic-sized tensor).

The following is the sample code using one javascript code sentence.

Pre Authentication Login
${prefs}= Create Dictionary
… credentials_enable_service=${FALSE}
… profile.password_manager_enabled=${FALSE}

${options}=    Evaluate    sys.modules['selenium.webdriver'].ChromeOptions()    sys
Call Method    ${options}    add_experimental_option    prefs    ${prefs}
Call Method    ${options}    add_experimental_option    detach    ${TRUE}

Open Browser    ${URL}    browser=${BROWSER}    options=${options}

# Try to use Selenium directly
Execute JavaScript    window.navigator.credentials = { get: () => Promise.resolve({ id: '${preauth_username}', password: '${preauth_password}' }) };

Go To    ${URL}

Let me request you guys’ helps on this.

With Regards,
Zuzan

2

Hi Zuzan,

The error message you describe, seems to be related to “TensorFlow” and from what I know, this is not normally required for SeleniumLibrary.

I’ll suggest trying to narrow down the problem a little:

  • Try running Open Browser with the URL about:blank so the browser just opens with a blank page, this will confirm that Selenium is not causing this issue
  • Next using Go To to navigate away from the blank page to the user you specified with ${URL} if the error occurs here, then you’ve narrowed it down to either the page itself generating the error or something that the page required that’s causing it (a plug-in maybe?)

I’ve never seen this error myself but hopefully this helps lead to the answer :crossed_fingers:

Also while I’m no expert on tensor related stuff, from what I know they normally require graphics cards and GPU’s more than CPU’s. where are you running this code? is it on a physical machine, virtual machine, docker container or somewhere else? I ask because if your tensor code without a GPU and appropriate GPU drivers that could be related to the error you’re getting?

Hope this helps,

Dave.