thd format is not supported with hierarchical CP implementation

[stormchasingg]

Describe the bug
What the bug is, and how to reproduce, better with screenshots(描述bug以及复现过程，最好有截图)
add parameters "hierarchical_context_parallel_sizes":[2,2], "cp_comm_type": "a2a+p2p", error report:

[rank0]: Traceback (most recent call last):
[rank0]:   File "/ms-swift/swift/cli/_megatron/sft.py", line 5, in <module>
[rank0]:     megatron_sft_main()
[rank0]:   File "/ms-swift/swift/megatron/train/sft.py", line 79, in megatron_sft_main
[rank0]:     return MegatronSft(args).main()
[rank0]:   File "/ms-swift/swift/llm/base.py", line 49, in main
[rank0]:     result = self.run()
[rank0]:   File "/ms-swift/swift/megatron/train/sft.py", line 69, in run
[rank0]:     self.trainer.train(train_dataset, val_dataset, data_collator)
[rank0]:   File "/ms-swift/swift/megatron/trainers/base.py", line 749, in train
[rank0]:     pretrain(
[rank0]:   File "/Megatron-LM/megatron/training/training.py", line 864, in pretrain
[rank0]:     iteration, num_floating_point_operations_so_far = train(
[rank0]:   File "/Megatron-LM/megatron/training/training.py", line 2279, in train
[rank0]:     ) = train_step(
[rank0]:   File "/ms-swift/swift/megatron/trainers/base.py", line 302, in train_step
[rank0]:     return self._origin_train_step(forward_step_func, new_data_iterator, model, optimizer, opt_param_scheduler,
[rank0]:   File "/Megatron-LM/megatron/training/training.py", line 1395, in train_step
[rank0]:     losses_reduced = forward_backward_func(
[rank0]:   File "/Megatron-LM/megatron/core/pipeline_parallel/schedules.py", line 1889, in forward_backward_pipelining_without_interleaving
[rank0]:     output_tensor, num_tokens = forward_step(
[rank0]:   File "/Megatron-LM/megatron/core/pipeline_parallel/schedules.py", line 289, in forward_step
[rank0]:     output_tensor, loss_func = forward_step_func(data_iterator, model)
[rank0]:   File "/ms-swift/swift/megatron/trainers/trainer.py", line 116, in forward_step
[rank0]:     output_tensor = model(**data)
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
[rank0]:     return self._call_impl(*args, **kwargs)
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
[rank0]:     return forward_call(*args, **kwargs)
[rank0]:   File "/Megatron-LM/megatron/core/distributed/data_parallel_base.py", line 22, in forward
[rank0]:     return self.module(*inputs, **kwargs)
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
[rank0]:     return self._call_impl(*args, **kwargs)
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
[rank0]:     return forward_call(*args, **kwargs)
[rank0]:   File "/Megatron-LM/megatron/core/transformer/module.py", line 237, in forward
[rank0]:     outputs = self.module(*inputs, **kwargs)
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
[rank0]:     return self._call_impl(*args, **kwargs)
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
[rank0]:     return forward_call(*args, **kwargs)
[rank0]:   File "/ms-swift/swift/megatron/model/gpt_model.py", line 196, in forward
[rank0]:     hidden_states = self.decoder(
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
[rank0]:     return self._call_impl(*args, **kwargs)
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
[rank0]:     return forward_call(*args, **kwargs)
[rank0]:   File "/Megatron-LM/megatron/core/transformer/transformer_block.py", line 563, in forward
[rank0]:     hidden_states = self._checkpointed_forward(
[rank0]:   File "/Megatron-LM/megatron/core/transformer/transformer_block.py", line 454, in _checkpointed_forward
[rank0]:     hidden_states, context = checkpoint_handler(custom(layer_idx, layer_idx + 1))
[rank0]:   File "/Megatron-LM/megatron/core/transformer/transformer_block.py", line 417, in checkpoint_handler
[rank0]:     return tensor_parallel.checkpoint(
[rank0]:   File "/Megatron-LM/megatron/core/tensor_parallel/random.py", line 477, in checkpoint
[rank0]:     return CheckpointFunction.apply(function, distribute_saved_activations, *args)
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/torch/autograd/function.py", line 575, in apply
[rank0]:     return super().apply(*args, **kwargs)  # type: ignore[misc]
[rank0]:   File "/Megatron-LM/megatron/core/tensor_parallel/random.py", line 423, in forward
[rank0]:     outputs = run_function(*args)
[rank0]:   File "/Megatron-LM/megatron/core/transformer/transformer_block.py", line 388, in custom_forward
[rank0]:     hidden_states, context = layer(
[rank0]:   File "/Megatron-LM/megatron/core/transformer/transformer_layer.py", line 875, in __call__
[rank0]:     return super(MegatronModule, self).__call__(*args, **kwargs)
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
[rank0]:     return self._call_impl(*args, **kwargs)
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
[rank0]:     return forward_call(*args, **kwargs)
[rank0]:   File "/ms-swift/swift/megatron/init.py", line 434, in forward
[rank0]:     hidden_states, context = self._forward_attention(*_args, **kwargs)
[rank0]:   File "/Megatron-LM/megatron/core/transformer/transformer_layer.py", line 501, in _forward_attention
[rank0]:     attention_output_with_bias = self.self_attention(
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
[rank0]:     return self._call_impl(*args, **kwargs)
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
[rank0]:     return forward_call(*args, **kwargs)
[rank0]:   File "/Megatron-LM/megatron/core/transformer/attention.py", line 737, in forward
[rank0]:     core_attn_out = self.core_attention(
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
[rank0]:     return self._call_impl(*args, **kwargs)
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
[rank0]:     return forward_call(*args, **kwargs)
[rank0]:   File "/Megatron-LM/megatron/core/extensions/transformer_engine.py", line 931, in forward
[rank0]:     core_attn_out = super().forward(
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/transformer_engine/pytorch/attention/dot_product_attention/dot_product_attention.py", line 1026, in forward
[rank0]:     return self.flash_attention(
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
[rank0]:     return self._call_impl(*args, **kwargs)
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
[rank0]:     return forward_call(*args, **kwargs)
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/transformer_engine/pytorch/attention/dot_product_attention/backends.py", line 659, in forward
[rank0]:     output = attn_forward_func_with_cp(
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/transformer_engine/pytorch/attention/dot_product_attention/context_parallel.py", line 3619, in attn_forward_func_with_cp
[rank0]:     out = AttnFuncWithCPAndKVP2P.apply(*args)
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/torch/autograd/function.py", line 575, in apply
[rank0]:     return super().apply(*args, **kwargs)  # type: ignore[misc]
[rank0]:   File "/opt/conda/lib/python3.10/site-packages/transformer_engine/pytorch/attention/dot_product_attention/context_parallel.py", line 469, in forward
[rank0]:     qkv_format != "thd"
[rank0]: AssertionError: thd format is not supported with hierarchical CP implementation yet!

Your hardware and system info
Write your system info like CUDA version/system/GPU/torch version here(在这里给出硬件信息和系统信息，如CUDA版本，系统，GPU型号和torch版本等)
platform H800
pytorch 2.7
megatron-lm branch core_r0.13.0
transformer_engine 2.4.0

Additional context
Add any other context about the problem here(在这里补充其他信息)