Neville Li
Neville Li
We need versioning to prevent loading settings from incompatible versions of featran
I can reproduce with the following code on scala 2.12. ``` import java.io.{ByteArrayOutputStream, ObjectOutputStream} import org.scalacheck._ object SerDe { def main(args: Array[String]): Unit = { val oos = new ObjectOutputStream(new...
- [ ] filter/map before groupBy/join - [ ] sum/aggregate vs groupBy + reduce values - [x] hashJoins, side inputs, broadcasts - [x] scala collection tricks - [x] java &...
Sparse vectors are commonly used to extract features, and we should make sure it works seamlessly, let's at least add tests for both models (and maybe example).
We should add some convenience methods, maybe in `TensorFlowExtras`, to feed Featran `FloatSparseArray` and `DoubleSparseArray` into TensorFlow. https://github.com/spotify/featran/blob/master/java/src/main/scala/com/spotify/featran/java/JavaOps.scala#L126 The feeding code might look like this: ```java runner .feed("input/raw_indices", Tensors.create(new long[]{0,...
What and how do we want to expose metadata about a model and feature spec/settings: * how do squads want to consume it? * what is this information gonna be...
Similar to the ones in featran: https://github.com/spotify/featran/tree/master/jmh/src/test/scala/com/spotify/featran/jmh
We're using some eager/expensive collections like `List` in semigroup types (the `B` in `Aggregator[A, B, C]`) and should probably use lazier equivalent instead.
## Description Rename code references of Presto to Trino per https://github.com/trinodb/trino/pull/14149#discussion_r974162083 - Only code references, e.g. method & variable names, comments - String literals & config files are kept as...
`MountableFile.transferTo` creates an empty file in the container when the source file doesn't exist. This change fixes that.