optimized funcion for cpu usages

app.py

@@ -9,7 +9,7 @@ app = FastAPI()
 # Load the DeepSeek model and tokenizer
 MODEL_NAME = "deepseek-ai/deepseek-coder-1.3b-instruct"
 tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
-model = AutoModelForCausalLM.from_pretrained(MODEL_NAME, torch_dtype=torch.float16).to("cpu")  # Use "cuda" if available
+model = AutoModelForCausalLM.from_pretrained(MODEL_NAME, torch_dtype=torch.float32).to("cpu")
 
 
 class ChatRequest(BaseModel):
@@ -19,9 +19,10 @@ def generate_sql_query(user_input: str) -> str:
     """
     Generate an SQL query from a natural language query using the DeepSeek model.
     """
-    inputs = tokenizer(user_input, return_tensors="pt")
-    outputs = model.generate(**inputs, max_length=400)
+    inputs = tokenizer(user_input, return_tensors="pt", padding="longest", truncation=True)
+    outputs = model.generate(**inputs, max_length=400, do_sample=False, num_beams=1)
     sql_query = tokenizer.decode(outputs[0], skip_special_tokens=True)
+
     return sql_query