[MRG] Benchmark models for 4.0

gensim/models/word2vec.py

@@ -1638,7 +1638,7 @@ def _log_train_end(self, raw_word_count, trained_word_count, total_elapsed, job_
 
         """
         logger.info(
-            "training on a %i raw words (%i effective words) took %.1fs, %.0f effective words/s",
+            "training on %i raw words (%i effective words) took %.1fs, %.0f effective words/s",
             raw_word_count, trained_word_count, total_elapsed, trained_word_count / total_elapsed
         )
 

gensim/scripts/benchmark.py

@@ -0,0 +1,35 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+#
+# Copyright (C) 2020 Radim Rehurek <[email protected]>
+
+"""
+Help script (template) for benchmarking. Run with:
+
+  /usr/bin/time --format "%E elapsed\n%Mk max mem" python -m gensim.scripts.benchmark ~/gensim-data/text9/text9.txt
+
+"""
+
+import logging
+import sys
+
+from gensim.models.word2vec import Text8Corpus, LineSentence  # noqa: F401
+from gensim.models import FastText, Word2Vec, Doc2Vec, Phrases  # noqa: F401
+from gensim import __version__
+
+logger = logging.getLogger(__name__)
+
+
+if __name__ == "__main__":
+    logging.basicConfig(
+        format='%(asctime)s [%(processName)s/%(process)d] [%(levelname)s] %(name)s:%(lineno)d: %(message)s',
+        level=logging.INFO,
+    )
+
+    if len(sys.argv) < 2:
+        print(globals()['__doc__'] % locals())
+        sys.exit(1)
+
+    corpus = Text8Corpus(sys.argv[1])  # text8/text9 format from http://mattmahoney.net/dc/textdata.html
+    cls = FastText
+    cls(corpus, workers=12, epochs=1).save(f'/tmp/{cls.__name__}.gensim{__version__}')

gensim/utils.py

@@ -535,7 +535,7 @@ def _smart_save(self, fname, separately=None, sep_limit=10 * 1024**2, ignore=fro
         in separate files. The automatic check is not performed in this case.
 
         """
-        logger.info("saving %s object under %s, separately %s", self.__class__.__name__, fname, separately)
+        logger.info("saving %s under %s, separately %s", self, fname, separately)
 
         compress, subname = SaveLoad._adapt_by_suffix(fname)