Move tensors to GPU

handler.py

@@ -25,6 +25,7 @@ class EndpointHandler():
         """
         sentences = data.pop("inputs",data)
         encoded_input = self.tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')
+        encoded_input = {key: value.to(self.device) for key, value in encoded_input.items()}
         # Compute token embeddings
         with torch.no_grad():
             model_output = self.model(**encoded_input)