Update Bert encoder and vocab paths

Author: deliahu

examples/text-generator/encoder.py

@@ -105,11 +105,9 @@ def decode(self, tokens):
 
 def get_encoder():
     s3 = boto3.client("s3")
-    encoder = json.load(
-        s3.get_object(Bucket="cortex-examples", Key="gpt-2/124/encoder.json")["Body"]
-    )
+    encoder = json.load(s3.get_object(Bucket="cortex-examples", Key="gpt-2/encoder.json")["Body"])
     bpe_data = (
-        s3.get_object(Bucket="cortex-examples", Key="gpt-2/124/vocab.bpe")["Body"]
+        s3.get_object(Bucket="cortex-examples", Key="gpt-2/vocab.bpe")["Body"]
         .read()
         .decode("utf-8")
     )