Update README.md

README.md

@@ -1,3 +1,22 @@
----
-license: mit
----
+# ModelsNumbers
+
+Generated by Faker, contains data for testing searching model numbers using vectorized model numbers. 
+This dataset does NOT include embeddings.  See other datasets for a smaller sample, and one with embeddings.
+
+Size: 50,000 entries.
+
+*Columns*: 
+- brand
+- model_number
+- model_name
+- year
+- randomdata: between 1000 and 2000. Append to model_number if the faked value is under 6 characters.
+- model_search: remove some characters (see below) from model_number. This used for creating embeddings.
+- model_search_embeddings: only in those datasets with 'embeddings' added to filename.
+
+ 
+ 
+
+---
+license: mit
+---