scala 3 api batch 2 #31
Conversation
| @@ -5,14 +5,38 @@ import scala.language.implicitConversions | |||
| import scala.reflect.ClassTag | |||
|
|
|||
| object PythonUnion { | |||
| def convert[X <: py.Any, Y <: py.Any, A <: PythonType[X], B <: PythonType[Y]](u: A | B)(implicit ev1: ClassTag[A], ev2: ClassTag[B]): py.|[X, Y] = | |||
|
|
|||
| def fromPyhtonTypeAndScalaTypeUnion[X <: py.Any, A <: PythonType[X], B <: Any](u: A | B)(implicit ev1: ClassTag[A], ev2: ClassTag[B]): py.|[B, X] = | |||
There was a problem hiding this comment.
This cannot be implicit conversion without some clever type bound on B because it collides with the one beneath fromPyhtonTypesUnion
dotty-tensorflow/src/main/scala/me/shadaj/scalapy/tensorflow/api/keras/backend/Backend.scala
Show resolved
Hide resolved
...y-tensorflow/src/main/scala/me/shadaj/scalapy/tensorflow/api/keras/layers/MaxPooling2D.scala
Show resolved
Hide resolved
dotty-tensorflow/src/main/scala/me/shadaj/scalapy/tensorflow/api/scalaUtils/PythonUnion.scala
Outdated
Show resolved
Hide resolved
dotty-tensorflow/src/main/scala/me/shadaj/scalapy/tensorflow/api/keras/backend/Backend.scala
Outdated
Show resolved
Hide resolved
...ensorflow/src/main/scala/me.shadaj.scalapy.tensorflow/example/BidirectionalLSTMExample.scala
Outdated
Show resolved
Hide resolved
|
cleared up git history |
| println(s"x_test shape: ${x_test1.shape}") | ||
| val xTrain1 = sequence.padSequences(xTrain, maxLen = Some(maxlen)).astype(np.float32) | ||
| val xTest1 = sequence.padSequences(xTest, maxLen = Some(maxlen)).astype(np.float32) | ||
| val yTrain1 = yTrain.astype(np.float32) |
There was a problem hiding this comment.
Why are you casting interger ({0, 1}?) value to a float?
There was a problem hiding this comment.
There are no generic types implemented in lower layer (scala 2) for now, we picked Float as a universal type because it was convenient. (created issue to add generic types VirtuslabRnD/scalapy#3 )
But we can use Numeric in this layer and cast it to Float when passing to scala 2 layer.
There was a problem hiding this comment.
Made an issue, will take care of that later, It causes build problems
dotty-tensorflow/src/main/scala/me/shadaj/scalapy/tensorflow/api/keras/models/Models.scala
Outdated
Show resolved
Hide resolved
| import preprocessing.Preprocessing | ||
| import models.Models | ||
| import datasets.Datasets | ||
| import backend.Backend | ||
| class Keras private[api] (val underlying: PyKeras) extends PythonType[PyKeras] with PythonModule { |
There was a problem hiding this comment.
This comment is for whole PR as well as for other batches -- I'm strongly convinced that exposing public api without a line of comment does not make users' lives easier.
There was a problem hiding this comment.
well thats true, although documentation can be found on tensorflow website.
Anyway we had somewhat inconclusive discussion about this. Lets make an issue #38 :D and don't worry for now.
dotty-tensorflow/src/main/scala/me/shadaj/scalapy/tensorflow/api/scalaUtils/PythonUnion.scala
Show resolved
Hide resolved
dotty-tensorflow/src/main/scala/me/shadaj/scalapy/tensorflow/api/keras/backend/Backend.scala
Outdated
Show resolved
Hide resolved
dotty-tensorflow/src/main/scala/me/shadaj/scalapy/tensorflow/api/keras/datasets/Datasets.scala
Outdated
Show resolved
Hide resolved
...ensorflow/src/main/scala/me.shadaj.scalapy.tensorflow/example/BidirectionalLSTMExample.scala
Outdated
Show resolved
Hide resolved
dotty-tensorflow/src/main/scala/me/shadaj/scalapy/tensorflow/api/keras/layers/LSTM.scala
Show resolved
Hide resolved
dotty-tensorflow/src/main/scala/me/shadaj/scalapy/tensorflow/api/keras/models/Sequential.scala
Outdated
Show resolved
Hide resolved
dotty-tensorflow/src/main/scala/me/shadaj/scalapy/tensorflow/api/keras/models/Sequential.scala
Outdated
Show resolved
Hide resolved
dotty-tensorflow/src/main/scala/me/shadaj/scalapy/tensorflow/api/keras/models/Sequential.scala
Outdated
Show resolved
Hide resolved
to be clear
enumswill be added later.