Hi, I have a few questions and I can't seem to fin...

# kotlindl

[Dawid Dębowski]

Hi, I have a few questions and I can't seem to find an answer on the Internet:
Some information for the beginning - we're creating a model that should predict whether OCR should be made using a cloud service or inside the application. Here are the questions:
1. Is it possible to create a dataset with multidimensional output? We hope the model will be able to both tell where the image should be processed, as well as it should tell how much time it will spend on processing image. I tried to use OnHeapDataset but labels have to be one-dimensional Float Array, so there's no way of doing even a data class.
2. Is it possible to do more training once the model was taught? What I mean is that we're going to train the model on the side with some data and we're wondering if it will be possible to do some additional training during the lifecycle of the application (it's an university project, so we don't care about performance of the solution as much). Will another call of fit function suffice?
Thank you for the answers!

[Julia Beliaeva]

Is it possible to create a dataset with multidimensional output?

Unfortunately, no. Current dataset api only works with Float output.

Will another call of fit function suffice?

Yes, you can call fit another time to get perform training. Take a look at this example: https://github.com/Kotlin/kotlindl/blob/master/examples/src/main/kotlin/examples/i[…]ptimizers/LeNetMnistExportImportToJSONWithAdamOptimizerState.kt

[Dawid Dębowski]

Okok, thank you so much ❤️

[Dawid Dębowski]

Hi, follow up question - is there a way to build a model from scratch in Android (and call the fit method on this model)?

[Julia Beliaeva]

Hi, no, it's not possible right now. You'll need to train your model elsewhere and convert it to onnx format to use with KotlinDL in Android.

[Dawid Dębowski]

okok, thanks