[bugfix]Qwen2.5VL accurate question (#6975)

### What this PR does / why we need it?
The attention mechanism in the ViT model architecture of Qwen2.5VL
consists of two parts and does not support using cache to pass sequence
lengths.
### Does this PR introduce _any_ user-facing change?
remove seq_lens_cache
### How was this patch tested?

- vLLM version: v0.16.0
- vLLM main:
15d76f74e2

---------

Signed-off-by: tanhaoan333 <tanhaoan@huawei.com>

vllm_ascend/ops/mm_encoder_attention.py

@@ -62,7 +62,6 @@ class AscendMMEncoderAttention(MMEncoderAttention):
             prefix=prefix,
         )
 
-        self.layer_index = int("".join(filter(str.isdigit, prefix)))
         self.enable_pad = self.head_size > MIN_PAD_SIZE and self.head_size < MAX_PAD_SIZE
         self.scale_value = self.head_size**-0.5
 
@@ -103,12 +102,9 @@ class AscendMMEncoderAttention(MMEncoderAttention):
         is_reshaped = query.dim() == 4
 
         # Directly use seq_lens cpu cache to avoid d2h copy.
-        global seq_lens_cpu_cache
-        if self.layer_index == 0:
-            if cu_seqlens is None:
-                cu_seqlens = torch.arange(0, (bsz + 1) * q_len, step=q_len, dtype=torch.int32, device="cpu")
-            # Update seq_lens cpu cache.
-            seq_lens_cpu_cache = torch.diff(cu_seqlens).to("cpu")
+        if cu_seqlens is None:
+            cu_seqlens = torch.arange(0, (bsz + 1) * q_len, step=q_len, dtype=torch.int32, device="cpu")
+        seq_lens_cpu = torch.diff(cu_seqlens).to("cpu")
 
         # q, k, v: [b, s, head, head_dim] -> [b * s, head, head_dim]
         q, k, v = self._reshape_qkv_to_3d(query, key, value, bsz, q_len, kv_len)
@@ -128,7 +124,7 @@ class AscendMMEncoderAttention(MMEncoderAttention):
             query=q,
             key=k,
             value=v,
-            seq_len=seq_lens_cpu_cache,
+            seq_len=seq_lens_cpu,
             scale_value=self.scale_value,
             num_heads=self.num_heads,
             num_kv_heads=self.num_kv_heads,