refactor: switch to huggingface_hub InferenceClient for Qwen, remove local transformers usage

agent.py

@@ -20,7 +20,7 @@ from typing_extensions import TypedDict
 from pydantic import BaseModel, Field
 
 # LangChain HuggingFace Integration
-from transformers import AutoModelForCausalLM, AutoTokenizer
+from huggingface_hub import InferenceClient
 
 from utils import (
     process_question_with_tools,
@@ -55,35 +55,27 @@ class AIBrain:
     def __init__(self):
         self.model_name = "Qwen/Qwen3-8B"
         
-        print("🧠 Initializing Qwen3-8B với transformers gốc...")
-        self.tokenizer = AutoTokenizer.from_pretrained(self.model_name)
-        self.model = AutoModelForCausalLM.from_pretrained(
-            self.model_name,
-            torch_dtype="auto",
-            device_map="auto"
+        print("🧠 Initializing Qwen3-8B với huggingface_hub InferenceClient...")
+        self.client = InferenceClient(
+            provider="auto",
+            api_key=os.environ["HF_TOKEN"],
         )
-        print("✅ Qwen3 AI Brain với transformers đã sẵn sàng")
+        print("✅ Qwen3 AI Brain với huggingface_hub InferenceClient đã sẵn sàng")
     
     def _generate_with_qwen3(self, prompt: str, max_tokens: int = 2048) -> str:
-        """Sinh text với Qwen3 bằng transformers gốc, thinking mode tắt"""
+        """Sinh text với Qwen3 bằng huggingface_hub InferenceClient"""
         try:
-            messages = [{"role": "user", "content": prompt}]
-            text = self.tokenizer.apply_chat_template(
-                messages,
-                tokenize=False,
-                add_generation_prompt=True,
-                enable_thinking=False
+            messages = [
+                {"role": "user", "content": prompt}
+            ]
+            completion = self.client.chat.completions.create(
+                model=self.model_name,
+                messages=messages,
+                max_tokens=max_tokens
             )
-            model_inputs = self.tokenizer([text], return_tensors="pt").to(self.model.device)
-            generated_ids = self.model.generate(
-                **model_inputs,
-                max_new_tokens=max_tokens
-            )
-            output_ids = generated_ids[0][len(model_inputs.input_ids[0]):].tolist()
-            response = self.tokenizer.decode(output_ids, skip_special_tokens=True).strip("\n")
-            return response
+            return completion.choices[0].message.content
         except Exception as e:
-            print(f"⚠️ Qwen3 generation error: {str(e)}")
+            print(f"⚠️ Qwen3 InferenceClient error: {str(e)}")
             return f"AI generation failed: {str(e)}"
     
     def analyze_question(self, question: str, task_id: str = "") -> Dict[str, Any]: