updated model config and added logs

main.py

@@ -85,20 +85,31 @@ def run_ai_inference(diff: str) -> str:
     if not model or not tokenizer:
         raise RuntimeError("Model is not loaded.")
 
+    # Simplified, shorter prompt for faster inference
     messages = [
         {
             "role": "system",
-            "content": """You are an expert code reviewer. Your task is to analyze a pull request diff and provide constructive feedback.\nAnalyze the provided diff and identify potential issues, suggest improvements, or point out good practices.\n\nIMPORTANT: Respond with a JSON array of comment objects. Each object must have three fields: 'file_path', 'line_number', and 'comment_text'.\nThe 'file_path' should be the full path of the file being changed.\nThe 'line_number' must be an integer corresponding to the line number in the *new* version of the file where the comment applies.\nThe 'comment_text' should be your concise and clear review comment.\n\nExample response format:\n[\n    {\n        "file_path": "src/utils/helpers.py",\n        "line_number": 42,\n        "comment_text": "This function could be simplified by using a list comprehension."\n    }\n]\n\nDo not add any introductory text or explanations outside of the JSON array.\n"""
+            "content": "You are a code reviewer. Analyze the diff and return a JSON array of comments with file_path, line_number, and comment_text fields. Be concise."
         },
         {
             "role": "user",
-            "content": f"Here is the diff to review:\n\n```diff\n{diff}\n```"
+            "content": f"Review this diff:\n{diff[:1000]}"  # Limit diff size for faster processing
         }
     ]
 
     inputs = tokenizer.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt")
     
-    outputs = model.generate(inputs, max_new_tokens=1024, do_sample=False, top_k=50, num_return_sequences=1, eos_token_id=tokenizer.eos_token_id)
+    # Optimized generation parameters for speed
+    outputs = model.generate(
+        inputs, 
+        max_new_tokens=256,  # Reduced from 1024
+        do_sample=False,
+        temperature=0.1,     # Lower temperature for more focused output
+        num_return_sequences=1, 
+        eos_token_id=tokenizer.eos_token_id,
+        pad_token_id=tokenizer.eos_token_id,
+        use_cache=True       # Enable KV cache for faster generation
+    )
     
     response_text = tokenizer.decode(outputs[0][len(inputs[0]):], skip_special_tokens=True)
     return response_text.strip()
@@ -109,10 +120,15 @@ def parse_ai_response(response_text: str) -> list[ReviewComment]:
     """
     print(f"Raw AI Response:\n---\n{response_text}\n---")
     
-    json_match = re.search(r'\[.*\]', response_text, re.DOTALL)
+    json_match = re.search(r'\[.*?\]', response_text, re.DOTALL)
     if not json_match:
         print("Warning: Could not find a JSON array in the AI response.")
-        return []
+        # Return a simple fallback comment
+        return [ReviewComment(
+            file_path="unknown", 
+            line_number=1, 
+            comment_text="AI review completed - no specific issues found."
+        )]
 
     json_string = json_match.group(0)
     
@@ -123,7 +139,12 @@ def parse_ai_response(response_text: str) -> list[ReviewComment]:
     except (json.JSONDecodeError, TypeError, KeyError) as e:
         print(f"Error parsing JSON from AI response: {e}")
         print(f"Invalid JSON string: {json_string}")
-        return []
+        # Return a simple fallback comment
+        return [ReviewComment(
+            file_path="unknown", 
+            line_number=1, 
+            comment_text="AI review completed - response format issue."
+        )]
 
 # ----------------------------
 # 6. The API Endpoint
@@ -134,13 +155,23 @@ async def get_code_review(request: ReviewRequest):
     if not request.diff:
         raise HTTPException(status_code=400, detail="Diff content cannot be empty.")
 
+    import time
+    start_time = time.time()
+    print(f"Starting review request at {start_time}")
+
     try:
+        print("Running AI inference...")
         ai_response_text = run_ai_inference(request.diff)
+        print(f"AI inference completed in {time.time() - start_time:.2f} seconds")
+        
+        print("Parsing AI response...")
         parsed_comments = parse_ai_response(ai_response_text)
+        print(f"Total processing time: {time.time() - start_time:.2f} seconds")
+        
         return ReviewResponse(comments=parsed_comments)
 
     except Exception as e:
-        print(f"An unexpected error occurred: {e}")
+        print(f"An unexpected error occurred after {time.time() - start_time:.2f} seconds: {e}")
         raise HTTPException(status_code=500, detail="An internal error occurred while processing the review.")
 
 # ----------------------------