[BUG] Make sure the torch tensor is contiguous (#34)

* update bitblas

* Merge branch 'main' of https://github.com/microsoft/BitBLAS into main

* make sure weight is contiguous.

python/bitblas/ops/general_matmul.py

@@ -417,6 +417,7 @@ def transform_weight(self, weight, scale=None, zeros=None, bias=None):
         Returns:
         A list containing the transformed weight tensor and optionally the scale, zeros, and bias.
         """
+        weight = weight.contiguous()
         if self.W_dtype == self.A_dtype:
             if self.weight_transform is not None:
                 return self.weight_transform(weight.cpu()).cuda().contiguous()