Chirag Jain 50896ec574 Make nvcc threads configurable via environment variable (#885) 9 månader sedan
..
fused_softmax.cpp ed553e9238 Add Megatron attention implementation for benchmarking 2 år sedan
scaled_masked_softmax.h ed553e9238 Add Megatron attention implementation for benchmarking 2 år sedan
scaled_masked_softmax_cuda.cu ed553e9238 Add Megatron attention implementation for benchmarking 2 år sedan
scaled_upper_triang_masked_softmax.h ed553e9238 Add Megatron attention implementation for benchmarking 2 år sedan
scaled_upper_triang_masked_softmax_cuda.cu ed553e9238 Add Megatron attention implementation for benchmarking 2 år sedan
setup.py 50896ec574 Make nvcc threads configurable via environment variable (#885) 9 månader sedan
type_shim.h ed553e9238 Add Megatron attention implementation for benchmarking 2 år sedan