Refactor sentiment model initialization to use AutoModel and AutoTokenizer

model/sentiment.py

@@ -1,8 +1,11 @@
-from transformers import pipeline
+from transformers import pipeline, AutoModelForSequenceClassification, AutoTokenizer
 from typing import Any
 
 class Sentiment:
     def __init__(self, line: str) -> (list | list[Any] | Any | None):
-        self.pipe = pipeline("text-classification", model='./feel-it-italian-sentiment')
+        model = AutoModelForSequenceClassification.from_pretrained("./feel-it-italian-sentiment")
+        tokenizer = AutoTokenizer.from_pretrained("./feel-it-italian-sentiment")
+        
+        self.pipe = pipeline("text-classification", model=model, tokenizer=tokenizer)
         self.result = self.pipe(line)