Searched refs:layer_norm_grad_input_kernel_vectorized (Results 1 – 1 of 1) sorted by relevance
400 __global__ void layer_norm_grad_input_kernel_vectorized( in layer_norm_grad_input_kernel_vectorized() function1200 … layer_norm_grad_input_kernel_vectorized<<<blocks, num_threads(), nshared, cuda_stream>>>(dY_data, in LayerNormBackwardKernelImplInternal()