[Perf] Use value-only reduction for native per-token FP8 quantization (#59800)

Signed-off-by: jackLei0901 <841747318@qq.com>
Co-authored-by: Misha Goin <mgoin64@gmail.com>

vllm/model_executor/layers/quantization/input_quant_fp8.py

@@ -201,7 +201,7 @@ class QuantFP8(CustomOp):
 
         if scale is None:
             if self.group_shape == GroupShape.PER_TOKEN:
-                x_max, _ = x.abs().max(dim=-1)
+                x_max = x.abs().amax(dim=-1)
                 x_max = x_max.unsqueeze(-1).to(torch.float32)
                 if scale_ub is not None:
                     x_max = x_max.clamp(max=scale_ub)