Update sentiment_analyzer.py

sentiment_analyzer.py

@@ -1,6 +1,7 @@
 from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
 import torch
 import re
+import os # --- ADDED ---
 
 class NewsAnalyzer:
     def __init__(self, model_name="google/gemma-2-2b-it"):
@@ -12,10 +13,20 @@ class NewsAnalyzer:
         self.device = "cuda" if torch.cuda.is_available() else "cpu"
         print(f"Using device: {self.device}")
 
+        # --- ADDED: Get token from Space Secrets ---
+        hf_token = os.getenv("HF_TOKEN")
+        
+        if not hf_token:
+            print("Warning: HF_TOKEN secret not found. May fail to load gated models.")
+
         try:
-            self.tokenizer = AutoTokenizer.from_pretrained(model_name)
+            self.tokenizer = AutoTokenizer.from_pretrained(
+                model_name, 
+                token=hf_token # --- ADDED ---
+            )
             self.model = AutoModelForCausalLM.from_pretrained(
                 model_name,
+                token=hf_token, # --- ADDED ---
                 torch_dtype=torch.float16 if self.device == "cuda" else torch.float32,
                 device_map="auto" if self.device == "cuda" else None,
                 low_cpu_mem_usage=True
@@ -90,7 +101,6 @@ Reason: [Brief explanation of your analysis]"""
     def _parse_llm_analysis(self, response):
         """
         แยก sentiment, score, theme, impact และ explanation จาก LLM response
-        (เวอร์ชันแก้ไขให้ทนทานต่อ Markdown และข้อผิดพลาด)
         """
         sentiment = "Neutral"
         score = 0.5
@@ -99,9 +109,6 @@ Reason: [Brief explanation of your analysis]"""
         explanation = "Unable to parse"
 
         try:
-            # --- MODIFIED: Made Regex more robust ---
-            # (Handles optional markdown "**" and optional colon ":")
-            
             sentiment_line = re.search(r'\**Sentiment:?\**\s*(\w+)', response, re.IGNORECASE)
             if sentiment_line:
                 sentiment = sentiment_line.group(1).capitalize()
@@ -122,7 +129,6 @@ Reason: [Brief explanation of your analysis]"""
             reason_match = re.search(r'\**Reason:?\**\s*(.*)', response, re.DOTALL | re.IGNORECASE)
             if reason_match:
                 explanation = reason_match.group(1).strip()
-            # --- End of MODIFIED block ---
 
             if sentiment not in ["Positive", "Negative", "Neutral"]:
                 sentiment = "Neutral"