Fix grouped conv bwd wei dl kernel

92d725cf · Bartlomiej Kocot · 7303b516 · 92d725cf
Commit 92d725cf authored Oct 12, 2023 by Bartlomiej Kocot
Show whitespace changes
Inline Side-by-side

Showing with 2 additions and 2 deletions

include/ck/tensor_operation/gpu/device/impl/device_grouped_conv_bwd_weight_dl.hpp ...ion/gpu/device/impl/device_grouped_conv_bwd_weight_dl.hpp +2 -2

No files found.
--- a/include/ck/tensor_operation/gpu/device/impl/device_grouped_conv_bwd_weight_dl.hpp
+++ b/include/ck/tensor_operation/gpu/device/impl/device_grouped_conv_bwd_weight_dl.hpp
@@ -927,7 +927,7 @@ struct DeviceGroupedConvBwdWeight_Dl : public DeviceGroupedConvBwdWeight<NDimSpa
        Block2CTileMap block_2_ctile_map_;
        // for computing batch offset
-        ComputePtrOffsetOfStridedBatch compute_ptr_offset_of_batch_;
+        ComputePtrOffsetOfStridedBatch<I0> compute_ptr_offset_of_batch_;
        // element-wise op
        OutElementwiseOperation a_element_op_;
@@ -999,7 +999,7 @@ struct DeviceGroupedConvBwdWeight_Dl : public DeviceGroupedConvBwdWeight<NDimSpa
                    remove_reference_t<DeviceOp::BGridDesc_B_K0_N0_N1_K1>,
                    remove_reference_t<DeviceOp::CGridDesc_M0_M10_M11_N0_N10_N11>,
                    remove_reference_t<DeviceOp::Block2CTileMap>,
-                    ComputePtrOffsetOfStridedBatch,
+                    ComputePtrOffsetOfStridedBatch<I0>,
                    has_main_loop,
                    has_double_loop>;