spacey

src/generate_queries_alternative.py

@@ -6,10 +6,20 @@ import uuid
 import chromadb
 import spacy
 import numpy as np
-nlp = spacy.load("en_core_web_sm")
 import os
 #os.environ["TOKENIZERS_PARALLELISM"] = "false"
 
+import spacy
+
+def get_nlp():
+    try:
+        return spacy.load("en_core_web_sm")
+    except OSError:
+        from spacy.cli import download
+        download("en_core_web_sm")
+        return spacy.load("en_core_web_sm")
+nlp = get_nlp()
+
 wikibase_properties_id = {'instance of': 'P2',
  'reference URL': 'P24',
  'start time': 'P15',