Refactor sliding window configuration to Transformers best practice (#21927)

Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>

vllm/model_executor/models/qwen2.py

@@ -49,6 +49,7 @@ from vllm.model_executor.model_loader.weight_utils import (
     default_weight_loader, maybe_remap_kv_scale_name)
 from vllm.model_executor.sampling_metadata import SamplingMetadata
 from vllm.sequence import IntermediateTensors
+from vllm.transformers_utils.config import is_interleaved
 
 from .interfaces import SupportsLoRA, SupportsPP
 from .utils import (AutoWeightsLoader, PPMissingLayer, extract_layer_index,
@@ -285,8 +286,7 @@ class Qwen2Model(nn.Module):
         quant_config = vllm_config.quant_config
 
         # TODO (@robertgshaw2): see if this can be moved out
-        if (cache_config.sliding_window is not None
-                and hasattr(config, "max_window_layers")):
+        if is_interleaved(vllm_config.model_config.hf_text_config):
             assert config.max_window_layers == config.num_hidden_layers, (
                 "Sliding window for some but all layers is not supported. "
                 "This model uses sliding window but `max_window_layers` = {} "