Merge branch 'layer-norm-kerne-include' into 'main'

THC/THCDeviceUtils.cuh -> ATen/cuda/DeviceUtils.cuh in fused layer norm

See merge request ADLR/megatron-lm!333

megatron/fused_kernels/layer_norm_cuda_kernel.cu

@@ -21,7 +21,7 @@
 #include "ATen/ATen.h"
 #include "ATen/AccumulateType.h"
 #include "ATen/cuda/CUDAContext.h"
-#include <THC/THCDeviceUtils.cuh>
+#include "ATen/cuda/DeviceUtils.cuh"
 
 #include <cuda.h>
 #include <cuda_runtime.h>