[Cleanup] Remove unused attn_metadata parameter from Proposer classes (#4862)

The `attn_metadata` is not used by any draft proposer, so we can remove
it.


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

---------

Signed-off-by: Jade Zheng <zheng.shoujian@outlook.com>

vllm_ascend/spec_decode/interface.py

@@ -48,7 +48,6 @@ class Proposer:
                            positions: torch.Tensor = None,
                            num_scheduled_tokens: int = 0,
                            hidden_states: torch.Tensor = None,
-                           attn_metadata=None,
                            aux_hidden_states: torch.Tensor = None):
         """Called by execute_model in model_runner"""
         raise NotImplementedError