| .. |
|
attention
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
core
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
cpu
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
cutlass_extensions
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
mamba/mamba_ssm
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
moe
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
prepare_inputs
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
quantization
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
quickreduce
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
rocm
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
sparse/cutlass
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
activation_kernels.cu
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
cache.h
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
cache_kernels.cu
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
cuda_compat.h
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
cuda_utils.h
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
cuda_utils_kernels.cu
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
cuda_view.cu
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
cumem_allocator.cpp
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
custom_all_reduce.cu
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
custom_all_reduce.cuh
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
custom_all_reduce_test.cu
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
custom_quickreduce.cu
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
dispatch_utils.h
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
layernorm_kernels.cu
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
layernorm_quant_kernels.cu
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
ops.h
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
permute_cols.cu
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
pos_encoding_kernels.cu
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
sampler.cu
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
torch_bindings.cpp
|
first commit
|
2025-08-03 20:28:19 +08:00 |
|
type_convert.cuh
|
first commit
|
2025-08-03 20:28:19 +08:00 |