Hello, RNN Data Handler

datahandler.py

class DataHandler:
    def __init__(self, string):
        self.string = string
        
        characters = np.sort(list(set(string)))
        self.num_characters = len(characters)
        
        self.char_to_idx = { ch : i for i, ch in enumerate(characters) }
        self.idx_to_char = { i : ch for ch, i in self.char_to_idx.items() }
        
        self._process()
        
    def _process(self):
        data_torch = torch.tensor([self.make_onehot(ele).data.numpy() for ele in self.string])

        self.X = data_torch[:-1].float()
        self.y = torch.argmax(data_torch[1:], dim=1).long()
        
    def make_onehot(self, char):
        return torch.eye(self.num_characters)[self.char_to_idx[char]].float()