Update README.md

README.md

@@ -73,4 +73,7 @@ I also suspect that state-of-the-art approaches using GPT 3.5/4 or PaLM may be h
 Ideally, I would use these models for training my classifier, however, this would incur significant expense.
 Finally, the classifier can be tricked with some simple modification of LLM-written text, by adding rare words or long/short sentences.
 
-Given these limitations, I would emphasize that this detector serves as only a loose guide and should not be used for cheating detection.
+Given these limitations, I would emphasize that this detector serves as only a loose guide and should not be used for cheating detection.
+
+# Code
+The model inference is run by app.py, and dependencies are in requirements.txt. The training code is in main.ipynb.