KeyedVectors.get_embedding_layer(train_embeddings = False) returns a trainable Embedding layer.

[vrmpx]

Description

KeyedVectors.get_embedding_layer does not set trainable when returning the Keras Embedding layer.

Steps/Code/Corpus to Reproduce

from gensim.models.keyedvectors import KeyedVectors

embeddings = KeyedVectors.load_word2vec_format("GoogleNews-vectors-negative300.bin", binary = True)
embedding_layer = embeddings.get_embedding_layer(train_embeddings=False)
print(embedding_layer.trainable)

Expected Results

False

Actual Results

True

Versions

Linux-4.4.0-93-generic-x86_64-with-debian-stretch-sid
Python 3.6.1 |Continuum Analytics, Inc.| (default, May 11 2017, 13:09:58)
[GCC 4.4.7 20120313 (Red Hat 4.4.7-1)]
NumPy 1.13.1
SciPy 0.19.1
gensim 2.3.0
FAST_VERSION 2

[zsef123]

keyedvectors.py : 835

layer = Embedding(input_dim=weights.shape[0], output_dim=weights.shape[1], weights=[weights])

get_embedding_layer function is not use train_embeddings parameter

[menshikh-iv]

Fixed in #1558