Commit dd2635f4 authored by letaoqin's avatar letaoqin
Browse files

remove nousing comments

parent 761d6d82
...@@ -1305,7 +1305,6 @@ struct GridwiseBatchedMultiheadAttentionBackward_Qloop_Xdl_CShuffle_V1 ...@@ -1305,7 +1305,6 @@ struct GridwiseBatchedMultiheadAttentionBackward_Qloop_Xdl_CShuffle_V1
"D0BlockTransferSrcScalarPerVector * NThreadClusterLengths <= NPerBlock"); "D0BlockTransferSrcScalarPerVector * NThreadClusterLengths <= NPerBlock");
__host__ __device__ static constexpr auto GetD0BlockGlobalDescriptor_M0_N0_M1_M2_N1_M3() __host__ __device__ static constexpr auto GetD0BlockGlobalDescriptor_M0_N0_M1_M2_N1_M3()
{ {
// B1 matrix in LDS memory, dst of blockwise copy
return make_naive_tensor_descriptor_packed( return make_naive_tensor_descriptor_packed(
make_tuple(I1, I1, I1, D0M1, Number<NPerBlock>{}, D0M2)); make_tuple(I1, I1, I1, D0M1, Number<NPerBlock>{}, D0M2));
} }
......
...@@ -1361,7 +1361,6 @@ struct GridwiseBatchedMultiheadAttentionBackward_Qloop_Xdl_CShuffle_V2 ...@@ -1361,7 +1361,6 @@ struct GridwiseBatchedMultiheadAttentionBackward_Qloop_Xdl_CShuffle_V2
"D0BlockTransferSrcScalarPerVector * NThreadClusterLengths <= NPerBlock"); "D0BlockTransferSrcScalarPerVector * NThreadClusterLengths <= NPerBlock");
__host__ __device__ static constexpr auto GetD0BlockGlobalDescriptor_M0_N0_M1_M2_N1_M3() __host__ __device__ static constexpr auto GetD0BlockGlobalDescriptor_M0_N0_M1_M2_N1_M3()
{ {
// B1 matrix in LDS memory, dst of blockwise copy
return make_naive_tensor_descriptor_packed( return make_naive_tensor_descriptor_packed(
make_tuple(I1, I1, I1, D0M1, Number<NPerBlock>{}, D0M2)); make_tuple(I1, I1, I1, D0M1, Number<NPerBlock>{}, D0M2));
} }
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment