AlpinDale a113309876 kernel: add meta functions for ops to prevent graph breaks (#1019) 2 달 전
..
all_reduce f1d0b77c92 [0.6.0] Release Candidate (#481) 6 달 전
attention 3bb0f07461 chore: rename `task_handler` to `worker` (#985) 2 달 전
backup 3bb0f07461 chore: rename `task_handler` to `worker` (#985) 2 달 전
core f1ea7711bd core: do not compile ScalarType for torch < 2.4.0 (#938) 2 달 전
cpu a113309876 kernel: add meta functions for ops to prevent graph breaks (#1019) 2 달 전
cutlass_extensions 93bc863591 feat: Machete Kernels for Hopper GPUs (#842) 3 달 전
hadamard 5d288aa76c feat: add fast hadamard transformation kernels (#232) 1 년 전
mamba 0256ed236b feat: windows support (#790) 4 달 전
moe 9f3e7c86e2 feat: add fused Marlin MoE kernel (#934) 2 달 전
prepare_inputs f1d0b77c92 [0.6.0] Release Candidate (#481) 6 달 전
punica f1d0b77c92 [0.6.0] Release Candidate (#481) 6 달 전
quantization a113309876 kernel: add meta functions for ops to prevent graph breaks (#1019) 2 달 전
sampling bfc8988116 feat: add cuda sampling kernels for top_k and top_p (#828) 3 달 전
activation_kernels.cu f1d0b77c92 [0.6.0] Release Candidate (#481) 6 달 전
cache.h f1d0b77c92 [0.6.0] Release Candidate (#481) 6 달 전
cache_kernels.cu 0256ed236b feat: windows support (#790) 4 달 전
cuda_compat.h f1d0b77c92 [0.6.0] Release Candidate (#481) 6 달 전
cuda_utils.h 93bc863591 feat: Machete Kernels for Hopper GPUs (#842) 3 달 전
cuda_utils_kernels.cu f1d0b77c92 [0.6.0] Release Candidate (#481) 6 달 전
dispatch_utils.h f1d0b77c92 [0.6.0] Release Candidate (#481) 6 달 전
layernorm_kernels.cu e14223dce5 kernel: use `cub::BlockReduce` instead of custom impl (#895) 2 달 전
ops.h 93bc863591 feat: Machete Kernels for Hopper GPUs (#842) 3 달 전
permute_cols.cu 93bc863591 feat: Machete Kernels for Hopper GPUs (#842) 3 달 전
pos_encoding_kernels.cu f1d0b77c92 [0.6.0] Release Candidate (#481) 6 달 전
torch_bindings.cpp a113309876 kernel: add meta functions for ops to prevent graph breaks (#1019) 2 달 전