i want to build face recongination that builds in cousera course from starting so where to start i run classroom code in my local enviroment but it gives error so where to start i am confuse
Hi @Deepak11
Your question is a bit unclear!!!
you wants to create facial recognition model in Coursera Course?
Can you share the screenshot of the error you are talking about.
Regards
DP
i want to build a face recong, for myself building and running on my data …so i copy and starts modifying code given in lab. when i runs in my local machine its through that error and i dont know how to tackle it ,this code works well on coursera lab but fails in my machine throughing that error
Did you download all the files required you to run the assignment on your local Jupyter environment, I mean the metadata(utils) file.
Also can I know when you mention your data, it is data outside of what was used in Coursera?
So the metadata or utils might not work on this.
You could probably share what kind of data you are using compared to the data used in the Coursera labs.
Regards
DP
yes i downloaded all the files in my local machine and then runs i think its may be an error occur due to different version . coursera runs on late version of python and tensorflow …i dont know i am right or not …
look i downloaded all files which provided by coursera
What about the kind of the data you are using need to compile with the metadata provided by the course. So did you go the utils files if that would require any modifications from your side to be used for your model??
Note that the online course notebook uses a pretty old version of TF:
TF version 2.3.0
It could well be that a saved model created for TF 2.3.0 might not be readable by a more current version of TF. In order to diagnose this, you have (of course) two choices:
- Try installing an older version of TF in your local environment.
- Try to find a saved version of FaceNet that is compatible with current TF.
Here’s a place to look for pretrained models provided by TF.