rebase main and fix

Signed-off-by: zzhx1 <zzh_201018@outlook.com>

Author: zzhx1

tests/ut/ops/test_vocab_parallel_embedding.py

@@ -227,5 +227,5 @@ def test_get_logits(self):
         lmhead.quant_method = self.mock_quant_method
         lmhead.quant_method.apply = self.mock_quant_method.apply
         hidden_state = torch.randn(1, self.org_num_embeddings)
-        processed_logits = processor._get_logits(hidden_state, lmhead)
+        processor._get_logits(hidden_state, lmhead)
         self.mock_quant_method.apply.assert_called_once()