public class GetEmbeddingsMeanRoute extends Object implements spark.Route
[4 bytes] [num words bytes] [rest]
for each input word,
embedding dimensions 1 byte is allocated the byte-encoded
(length of the with 0 or 1 to indicate mean vector
embedding vectors) that the respective
word was (not) in
the vocabulary
Where the embedding vector encoding follows the format described in
EmbeddingServerUtils.convertVectorToByteArray(double[]).| Constructor and Description |
|---|
GetEmbeddingsMeanRoute(de.julielab.ml.embeddings.spi.WordEmbedding embedding) |
public GetEmbeddingsMeanRoute(de.julielab.ml.embeddings.spi.WordEmbedding embedding)
Copyright © 2018 JULIE Lab, Germany. All rights reserved.