fix(sentiment_analyzer): fix type warning from pyright

@@ -142,6 +142,7 @@ class WeiboMultilingualSentimentAnalyzer:
         """Select the best available torch device."""
         """Select the best available torch device."""
         if not TORCH_AVAILABLE:
         if not TORCH_AVAILABLE:
             return None
             return None
+        assert torch is not None
         if torch.cuda.is_available():
         if torch.cuda.is_available():
             return torch.device("cuda")
             return torch.device("cuda")
         mps_backend = getattr(torch.backends, "mps", None)
         mps_backend = getattr(torch.backends, "mps", None)
@@ -177,6 +178,8 @@ class WeiboMultilingualSentimentAnalyzer:
 
 
         try:
         try:
             print("正在加载多语言情感分析模型...")
             print("正在加载多语言情感分析模型...")
+            assert AutoTokenizer is not None
+            assert AutoModelForSequenceClassification is not None
 
 
             # 使用多语言情感分析模型
             # 使用多语言情感分析模型
             model_name = "tabularisai/multilingual-sentiment-analysis"
             model_name = "tabularisai/multilingual-sentiment-analysis"
@@ -300,7 +303,7 @@ class WeiboMultilingualSentimentAnalyzer:
                     error_message="输入文本为空或无效内容",
                     error_message="输入文本为空或无效内容",
                     analysis_performed=False,
                     analysis_performed=False,
                 )
                 )
-
+            assert self.tokenizer is not None
             # 分词编码
             # 分词编码
             inputs = self.tokenizer(
             inputs = self.tokenizer(
                 processed_text,
                 processed_text,
@@ -314,11 +317,13 @@ class WeiboMultilingualSentimentAnalyzer:
             inputs = {k: v.to(self.device) for k, v in inputs.items()}
             inputs = {k: v.to(self.device) for k, v in inputs.items()}
 
 
             # 预测
             # 预测
+            assert torch is not None
+            assert self.model is not None
             with torch.no_grad():
             with torch.no_grad():
                 outputs = self.model(**inputs)
                 outputs = self.model(**inputs)
                 logits = outputs.logits
                 logits = outputs.logits
                 probabilities = torch.softmax(logits, dim=1)
                 probabilities = torch.softmax(logits, dim=1)
-                prediction = torch.argmax(probabilities, dim=1).item()
+                prediction = int(torch.argmax(probabilities, dim=1).item())
 
 
             # 构建结果
             # 构建结果
             confidence = probabilities[0][prediction].item()
             confidence = probabilities[0][prediction].item()