[EPLB][Bugfix] EPLB support fp/bf16 (#5531)

### What this PR does / why we need it?
EPLB support dtype of fp/bf16.
### Does this PR introduce _any_ user-facing change?

### How was this patch tested?
w8a8_dynamic Baseline:
| dataset | version | metric | mode | vllm-api-general-chat |
|----- | ----- | ----- | ----- | -----|
| aime2024 | 604a78 | accuracy | gen | 86.67 |

w8a8_dynamic eplb:
| dataset | version | metric | mode | vllm-api-general-chat |
|----- | ----- | ----- | ----- | -----|
| aime2024 | 604a78 | accuracy | gen | 86.67 |

The fp16 conversation is normal.
The fp16 test is in progress.

Baseline fp16
| dataset | version | metric | mode | vllm-api-general-chat |
|----- | ----- | ----- | ----- | -----|
| aime2024 | 604a78 | accuracy | gen | 86.67 |

eplb fp16
| dataset | version | metric | mode | vllm-api-general-chat |
|----- | ----- | ----- | ----- | -----|
| aime2024 | 604a78 | accuracy | gen | 83.33 |

- vLLM version: v0.13.0
- vLLM main:
45c1ca1ca1

Signed-off-by: shenchuxiaofugui <1311027364@qq.com>

tests/ut/eplb/adaptor/test_vllm_adaptor.py

@@ -0,0 +1,39 @@
+import unittest
+from unittest.mock import MagicMock, patch
+
+import torch
+
+from vllm_ascend.eplb.adaptor.vllm_adaptor import VllmEplbAdaptor
+from transformers import DeepseekV2Config
+
+
+class TestVllmAdaptor(unittest.TestCase):
+    def setUp(self):
+        n_routed_experts = 256
+        mock_model = MagicMock()
+        mock_model.model.named_parameters.return_value = dict()
+        config = DeepseekV2Config(n_routed_experts=n_routed_experts)
+        mock_model.config = config
+        mock_model.get_expert_map.return_value = [i for i in range(n_routed_experts)]
+        mock_model.get_log2phy_map.return_value = [i for i in range(n_routed_experts)]
+        self.model = mock_model
+
+        self.mock_rank = patch("vllm_ascend.eplb.adaptor.vllm_adaptor.dist.get_rank", return_value=0).start()
+        self.mock_size = patch("vllm_ascend.eplb.adaptor.vllm_adaptor.dist.get_world_size", return_value=4).start()
+
+    @patch("torch.empty_like", return_value=torch.zeros(16, 32))
+    def test_init_fp16(self, mock_func):
+        self.model.quant_config = None
+        VllmEplbAdaptor(self.model)
+
+    @patch("torch.empty_like", return_value=torch.zeros(16, 32))
+    def test_init_w8a8(self, mock_func):
+        VllmEplbAdaptor(self.model)
+
+    def tearDown(self):
+        self.mock_rank.stop()
+        self.mock_size.stop()
+
+if __name__ == "__main__":
+    unittest.main()
+