diff --git a/examples/offline_inference_neuron.py b/examples/offline_inference_neuron.py
index 9392558a6fc00..c1009d50a778e 100644
--- a/examples/offline_inference_neuron.py
+++ b/examples/offline_inference_neuron.py
@@ -1,10 +1,14 @@
 import os
+
 from vllm import LLM, SamplingParams
 
+# Builds the cache for the neuron compiled model.
 os.environ['NEURONX_DUMP_TO'] = "./Cache"
-
+# creates XLA hlo graphs for all the context length buckets.
 os.environ['NEURON_CONTEXT_LENGTH_BUCKETS'] = "512,1024,2048"
+# creates XLA hlo graphs for all the token gen buckets.
 os.environ['NEURON_TOKEN_GEN_BUCKETS'] = "512,1024,2048"
+
 # Sample prompts.
 prompts = [
     "Hello, my name is",