There is no /models/model.tflite file available

larswaechter / quickdraw-cnn

A convolutional neural network using Tensorflow and Google's Quick, Draw! dataset to recognize hand drawn images.

https://larswaechter.dev/blog/recognizing-hand-drawn-doodles/

MIT License

11 stars 1 forks source link

There is no /models/model.tflite file available #3

Open abdul-rehman-2050 opened 9 months ago

abdul-rehman-2050 commented 9 months ago

Can you please link to the tflite model?

larswaechter commented 9 months ago

You can find it here :)

abdul-rehman-2050 commented 9 months ago

Ooh my bad. Thank you so much for this. This is a great work. I really appreciate this. Do we have keras model available as well? How much accuracy difference do we after converting to the tensorflow lite?

larswaechter commented 9 months ago

Unfortunately I don't have the Keras model any more, so you have to train it on your own. But the code is in the according Jupyter Notebook included. The Keras model has an accuracy of ~61%. I didn't calculate the accuracy after converting to TF Lite.

hyeonhh commented 1 week ago

Hi :) Can I use model.tflite file with android App?? @larswaechter