[TPU] Re-enable the Pallas MoE kernel (#18025)

Signed-off-by: Michael Goin <mgoin64@gmail.com>

vllm/model_executor/layers/fused_moe/layer.py

@@ -50,8 +50,7 @@ if is_rocm_aiter_moe_enabled():
 else:
     from vllm.model_executor.layers.fused_moe.fused_moe import grouped_topk
 if current_platform.is_tpu():
-    # the iterative moe implementation is used until the moe_pallas is fixed
-    from .moe_torch_iterative import fused_moe as fused_moe_pallas
+    from .moe_pallas import fused_moe as fused_moe_pallas
 else:
     fused_moe_pallas = None  # type: ignore
 logger = init_logger(__name__)