(self, data_idx)
| 813 | return {'text': np.array(tokens), "loss_mask": np.array(loss_mask)} |
| 814 | |
| 815 | def getidx(self, data_idx): |
| 816 | data = self.ds[data_idx] |
| 817 | tokens, loss_masks = data['tokens'], data['loss_masks'] |
| 818 | tokens = tokens + [self.tokenizer.get_command('eos').Id] |
| 819 | loss_masks = loss_masks + [1] |
| 820 | return tokens, loss_masks |
| 821 | |
| 822 | def pad_seq(self, seq, pad_id=None): |
| 823 | total_tokens = self.max_seq_len + 1 |
no test coverage detected