fix torchair mtp

Signed-off-by: MengqingCao <cmq0113@163.com>

Author: MengqingCao

tests/e2e/singlecard/spec_decode_v1/test_v1_mtp_torchair_correctness.py

@@ -92,7 +92,6 @@ def mtp_torchair_correctness(
     assert matches > int(0.66 * len(ref_outputs))
 
 
-@pytest.mark.skip("TODO: revert this skip")
 def test_mtp_torchair_correctness_piecewise(
     sampling_config: SamplingParams,
     model_name: str,

vllm_ascend/torchair/models/torchair_deepseek_mtp.py

@@ -23,6 +23,7 @@
 import torch.nn as nn
 from transformers import PretrainedConfig
 from vllm.attention.backends.abstract import AttentionMetadata
+from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, ModelConfig, VllmConfig
 from vllm.distributed import get_tensor_model_parallel_world_size
 from vllm.model_executor.layers.layernorm import RMSNorm
@@ -187,6 +188,7 @@ def compute_logits(
         return logits
 
 
+@support_torch_compile
 class TorchairDeepSeekMTP(DeepSeekMTP):
     # NOTE 1.The quantized MTP layer of deepseek on the NPU is not quantized;
     # NOTE 2.The description file generated by the current msmodelslim tool does not have