From 11e3f19764e5958247edcef4eb00110ef9a7fb8f Mon Sep 17 00:00:00 2001
From: ines <ines@ines.io>
Date: Wed, 25 Oct 2017 16:08:26 +0200
Subject: [PATCH] Fix vectors data added after training (see #1457)

---
 spacy/cli/train.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/spacy/cli/train.py b/spacy/cli/train.py
index 026b1fe445..da398751c3 100644
--- a/spacy/cli/train.py
+++ b/spacy/cli/train.py
@@ -146,8 +146,8 @@ def train(cmd, lang, output_dir, train_data, dev_data, n_iter=30, n_sents=0,
                 meta['accuracy'] = scorer.scores
                 meta['speed'] = {'nwords': nwords, 'cpu': cpu_wps,
                                  'gpu': gpu_wps}
-                meta['vectors'] = {'entries': nlp.vocab.vectors_length,
-                                   'width': 0}
+                meta['vectors'] = {'width': nlp.vocab.vectors_length,
+                                   'entries': len(nlp.vocab.vectors)}
                 meta['lang'] = nlp.lang
                 meta['pipeline'] = pipeline
                 meta['spacy_version'] = '>=%s' % about.__version__