Update main.py

main.py

@@ -12,40 +12,39 @@ n_gpu_layers = 0
 n_batch = 1024
 
 llm = LlamaCpp(
-    model_path="Phi-3-mini-4k-instruct-q4.gguf",  
+    model_path="Phi-3-mini-4k-instruct-q4.gguf",  # path to GGUF file
     temperature=0.1,
     n_gpu_layers=n_gpu_layers,
     n_batch=n_batch,
     verbose=True,
     n_ctx=4096
 )
+model0 = AutoModel.from_pretrained('sentence-transformers/paraphrase-TinyBERT-L6-v2')
 
-# Initializing sentence transformer model
 model = SentenceTransformer('sentence-transformers/paraphrase-TinyBERT-L6-v2')
 
 file_size = os.stat('Phi-3-mini-4k-instruct-q4.gguf')
-print("Model size: ", file_size.st_size, "bytes")
+print("model size ====> :", file_size.st_size, "bytes")
 
 
 @app.route('/cv', methods=['POST'])
 def get_skills():
     cv_body = request.json.get('cv_body')
 
+    # Simple inference example
     output = llm(
-        f"\n{cv_body}\nCan you list the skills mentioned in the CV?",
-        max_tokens=256,  
-        stop=[""], 
-        echo=True,  
+        f"<|user|>\n{cv_body}<|end|>\n<|assistant|>Can you list the skills mentioned in the CV?<|end|>",
+        max_tokens=256,  # Generate up to 256 tokens
+        stop=["<|end|>"], 
+        echo=True,  # Whether to echo the prompt
     )
 
     return jsonify({'skills': output})
 
-
 @app.get('/')
 def health():
     return jsonify({'status': 'Worked'})
 
-
 @app.route('/compare', methods=['POST'])
 def compare():
     employee_skills = request.json.get('employee_skills')