fix awq_dequantize (#4333)

2025-03-12 01:04:38 -07:00
parent 07f944631e
commit 0a3960f21f
1 changed files with 1 additions and 1 deletions
--- a/sgl-kernel/python/sgl_kernel/gemm.py
+++ b/sgl-kernel/python/sgl_kernel/gemm.py
@@ -7,7 +7,7 @@ from sgl_kernel.utils import _get_cache_buf, get_cuda_stream
 def awq_dequantize(
    qweight: torch.Tensor, scales: torch.Tensor, qzeros: torch.Tensor
 ) -> torch.ByteTensor:
-    return torch.ops.sgl_kernels.awq_dequantize(qweight, scales, qzeros)
+    return torch.ops.sgl_kernel.awq_dequantize(qweight, scales, qzeros)


 def int8_scaled_mm(mat_a, mat_b, scales_a, scales_b, out_dtype, bias=None):