[TPU] fix kv cache dtype in model runner (#19244)

Signed-off-by: Chengji Yao <chengjiyao@google.com>

Author: yaochengji

vllm/v1/worker/tpu_model_runner.py

@@ -29,7 +29,8 @@
                                     PlaceholderRange)
 from vllm.multimodal.utils import group_mm_inputs_by_modality
 from vllm.sequence import IntermediateTensors
-from vllm.utils import LayerBlockType, cdiv, is_pin_memory_available
+from vllm.utils import (STR_DTYPE_TO_TORCH_DTYPE, LayerBlockType, cdiv,
+                        is_pin_memory_available)
 from vllm.v1.attention.backends.pallas import (PallasAttentionBackend,
                                                PallasMetadata)
 from vllm.v1.core.encoder_cache_manager import compute_encoder_budget
@@ -138,6 +139,11 @@ def __init__(
 
         self.pin_memory = is_pin_memory_available()
         self.dtype = self.model_config.dtype
+        if cache_config.cache_dtype == "auto":
+            self.kv_cache_dtype = self.dtype
+        else:
+            self.kv_cache_dtype = STR_DTYPE_TO_TORCH_DTYPE[
+                cache_config.cache_dtype]
         self._hidden_states_dtype = self.dtype
 
         self.is_multimodal_model = model_config.is_multimodal_model
@@ -480,7 +486,7 @@ def get_kv_cache_spec(self) -> dict[str, KVCacheSpec]:
                         block_size=block_size,
                         num_kv_heads=attn_module.num_kv_heads,
                         head_size=attn_module.head_size,
-                        dtype=attn_module.dtype,
+                        dtype=self.kv_cache_dtype,
                         sliding_window=attn_module.sliding_window,
                         use_mla=False,
                     )
@@ -489,7 +495,7 @@ def get_kv_cache_spec(self) -> dict[str, KVCacheSpec]:
                         block_size=block_size,
                         num_kv_heads=attn_module.num_kv_heads,
                         head_size=attn_module.head_size,
-                        dtype=attn_module.dtype,
+                        dtype=self.kv_cache_dtype,
                         use_mla=False,
                     )
             elif attn_module.attn_type in (AttentionType.ENCODER,