[Model][1/N] Delete deepseek v2/v3 modeling codes. (#3189)

This PR deletes model codes of deepseek_v2 and deepseek_v3 to reuse the
model file from vLLM.

vLLM Ascend now uses custom ops register way instead of model file
hard-coding.

- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0

---------

Signed-off-by: whx-sjtu <2952154980@qq.com>

tests/ut/torchair/test_torchair_mla.py

@@ -525,6 +525,7 @@ class TestAscendMLATorchairImpl(TestBase):
             "v_head_dim": 128,
             "rotary_emb": MagicMock(),
             "q_proj": MagicMock(),
+            "q_b_proj": MagicMock(),
             "kv_b_proj": MagicMock(),
             "o_proj": MagicMock(),
             "kv_a_proj_with_mqa": MagicMock(),