Update app.py

app.py

@@ -1,10 +1,10 @@
 from transformers import AutoModelForCausalLM, AutoTokenizer
 from fastapi import FastAPI
 
-
+# Model ID on Hugging Face
 MODEL_ID = "rasyosef/Phi-1_5-Instruct-v0.1"
 
-
+# Load tokenizer and model from local cache (pre-downloaded in Docker build)
 tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
 model = AutoModelForCausalLM.from_pretrained(MODEL_ID)
 
@@ -16,7 +16,7 @@ def chat(query: str):
     GET /chat?query=Your+question
     Returns JSON: {"answer": "...model’s reply..."}
     """
-
+    # Build the instruction‐style prompt expected by Phi‐1.5 Instruct
     prompt = (
         "<|im_start|>system\nYou are a helpful assistant.<|im_end|>"
         "<|im_start|>user\n" + query + "<|im_end|>"
@@ -24,7 +24,7 @@ def chat(query: str):
     )
     inputs = tokenizer(prompt, return_tensors="pt")
     outputs = model.generate(**inputs, max_new_tokens=200)
-    
+    # Only decode newly generated tokens (skip the “prompt” tokens)
     response = tokenizer.decode(
         outputs[0][inputs.input_ids.shape[-1]:], 
         skip_special_tokens=True