WEBVTT

00:00.080 --> 00:01.040
Welcome back.

00:01.400 --> 00:05.320
We succeeded in getting the predicted MPG for the new car.

00:05.360 --> 00:10.720
Now let's export the model to Android Lite format.

00:10.760 --> 00:14.920
We are ready to export and use it in Android.

00:15.000 --> 00:23.360
So as we did before, let's start with converter from the TF TensorFlow Lite library.

00:23.600 --> 00:32.000
Dot Lite converter from Keras model and we pass that trained model.

00:32.240 --> 00:38.640
Then we need to use TF Lite model equals to converter dot.

00:38.640 --> 00:43.600
Convert and save the model using with open model.

00:43.720 --> 00:54.440
Let me name it as car mpg prediction model wb as f f dot write tf lite model.

00:54.440 --> 00:55.520
Let me run.

00:55.800 --> 00:57.320
And here we go.

00:57.600 --> 01:05.040
Let me see inside this function this folder if there is a file called car MPG prediction.

01:05.080 --> 01:06.080
Here we go.

01:06.120 --> 01:07.840
Congratulations guys.

01:07.880 --> 01:10.160
Everything worked fine.

01:10.480 --> 01:17.880
This is how we export the model to Android and you can download it to your PC.

01:18.160 --> 01:18.960
There we go.

01:19.000 --> 01:23.480
Let's use this model in Android Studio.