[Feat] Add custom Embedding tensor model parallel (#2616)

Similar to #2309 , this PR introduces Embedding tensor model parallel to
achieve decreasing of memory consumption. It support both eager mode and
graph mode.

And this PR refactor module tensor parallel configurations supported in
#2309, #2167, #2120, merge all config into `finegrained_tp_config` in
`additional_config`, including:
`lmhead_tensor_parallel_size`
`oproj_tensor_parallel_size`
`embedding_tensor_parallel_size`
`mlp_tensor_parallel_size`

- vLLM version: v0.12.0
- vLLM main:
ad32e3e19c

---------

Signed-off-by: zzhx1 <zzh_201018@outlook.com>
Signed-off-by: zzhxx <zhangzihang23@mails.ucas.ac.cn>
Co-authored-by: zzhx1 <zzh_201018@outlook.com>
Co-authored-by: chenxiao <Jaychou1620@Gmail.com>
Co-authored-by: zzhxx <zhangzihang23@mails.ucas.ac.cn>
Co-authored-by: Jade Zheng <zheng.shoujian@outlook.com>

vllm_ascend/envs.py

@@ -118,10 +118,6 @@ env_variables: Dict[str, Callable[[], Any]] = {
     # However, there might be hidden issues, and it is currently recommended to prioritize its use with dense models.
     "VLLM_ASCEND_ENABLE_DENSE_OPTIMIZE":
     lambda: bool(int(os.getenv("VLLM_ASCEND_ENABLE_DENSE_OPTIMIZE", '0'))),
-    # Whether to enable mlp optimize when tensor parallel is enabled.
-    # this feature in eager mode will get better performance.
-    "VLLM_ASCEND_ENABLE_MLP_OPTIMIZE":
-    lambda: bool(int(os.getenv("VLLM_ASCEND_ENABLE_MLP_OPTIMIZE", '0'))),
     # Whether to enable msMonitor tool to monitor the performance of vllm-ascend.
     "MSMONITOR_USE_DAEMON":
     lambda: bool(int(os.getenv("MSMONITOR_USE_DAEMON", '0'))),