[300I][Bugfix] fix unquant model weight nd2nz error (#6851)

### What this PR does / why we need it?
- This PR fixes an issue with weight format conversion for unquantized
models running on Ascend 310P devices.

- The changes refactor the logic for converting weights to the
FRACTAL_NZ format. Previously, this was handled in a 310P-specific
linear layer implementation (`AscendUnquantizedLinearMethod310`). This
implementation has been removed, and the logic is now centralized in the
`maybe_trans_nz` utility function. This function now checks if the
device is a 310P and applies the NZ format cast accordingly for
`float16`/`bfloat16` weights.

- This refactoring simplifies the code by removing platform-specific
duplication and ensures correct weight handling for unquantized models
on 310P.

### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
ut and local test
- vLLM version: v0.15.0
- vLLM main:
83b47f67b1

---------

Signed-off-by: Tflowers-0129 <2906339855@qq.com>

vllm_ascend/_310p/quantization/methods/w8a8_static.py

@@ -21,7 +21,7 @@ import torch
 import torch_npu
 
 from vllm_ascend.quantization.methods.base import AscendLinearScheme
-from vllm_ascend.utils import ACL_FORMAT_FRACTAL_NZ
+from vllm_ascend.utils import maybe_trans_nz
 
 from .registry import register_scheme
 
@@ -105,7 +105,7 @@ class AscendW8A8LinearMethod310(AscendLinearScheme):
         ).to(layer.aclnn_input_scale.dtype)
 
         # ---- matmul stage tensor ----
-        layer.weight.data = torch_npu.npu_format_cast(layer.weight.data, ACL_FORMAT_FRACTAL_NZ).transpose(0, 1)
+        layer.weight.data = maybe_trans_nz(layer.weight.data).transpose(0, 1)
 
         # ---- dequant stage tensors ----
         layer.weight_scale.data = torch.flatten(layer.weight_scale.data)

vllm_ascend/_310p/quantization/methods/w8a8s.py

@@ -21,7 +21,7 @@ import torch
 import torch_npu
 
 from vllm_ascend.quantization.methods.base import AscendLinearScheme
-from vllm_ascend.utils import ACL_FORMAT_FRACTAL_NZ
+from vllm_ascend.utils import maybe_trans_nz
 
 from .registry import register_scheme
 
@@ -84,4 +84,4 @@ class AscendW8A8SLinearMethod310(AscendLinearScheme):
         layer.aclnn_input_scale = layer.input_scale.data.repeat(expanding_factor)
         layer.aclnn_input_scale_reciprocal = 1.0 / layer.aclnn_input_scale.data
         layer.aclnn_input_offset = layer.input_offset.data.repeat(expanding_factor).to(layer.aclnn_input_scale.dtype)
-        layer.weight.data = torch_npu.npu_format_cast(layer.weight.data, ACL_FORMAT_FRACTAL_NZ)
+        layer.weight.data = maybe_trans_nz(layer.weight.data)