[Bugfix] Fix w8a8_int8 import error on NPU (#8147)

6737671c · Even Zhou · GitHub · fd63b62e · 6737671c
Unverified Commit 6737671c authored Jul 19, 2025 by Even Zhou Committed by GitHub Jul 18, 2025
Show whitespace changes
Inline Side-by-side

Showing with 2 additions and 0 deletions

python/sglang/srt/layers/quantization/w8a8_int8.py python/sglang/srt/layers/quantization/w8a8_int8.py +2 -0

No files found.
--- a/python/sglang/srt/layers/quantization/w8a8_int8.py
+++ b/python/sglang/srt/layers/quantization/w8a8_int8.py
@@ -754,6 +754,8 @@ class NPU_W8A8LinearMethod(LinearMethodBase):
        x: torch.Tensor,
        bias: Optional[torch.Tensor] = None,
    ) -> torch.Tensor:
+        from sglang.srt.layers.linear import RowParallelLinear
+
        if isinstance(layer, RowParallelLinear):
            tp_rank = get_tensor_model_parallel_rank()
            return self.quant_method.apply(layer, x, bias, tp_rank)