You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
ColossalAI/colossalai/kernel/cuda_native/csrc
encmps 79ccfa4310
[NFC] polish colossalai/kernel/cuda_native/csrc/multi_tensor_adam.cu code style (#667)
3 years ago
..
kernels [NFC] polish colossalai/kernel/cuda_native/csrc/kernels/include/cuda_util.h code style (#641) 3 years ago
colossal_C_frontend.cpp fix format (#568) 3 years ago
compat.h refactor kernel (#142) 3 years ago
cpu_adam.cpp [NFC] polish colossalai/kernel/cuda_native/csrc/cpu_adam.cpp code style (#636) 3 years ago
cpu_adam.h fix format (#608) 3 years ago
layer_norm_cuda.cpp [format]colossalai/kernel/cuda_native/csrc/layer_norm_cuda.cpp (#566) 3 years ago
layer_norm_cuda_kernel.cu [NFC] polish colossalai/kernel/cuda_native/csrc/layer_norm_cuda_kernel.cu code style (#661) 3 years ago
moe_cuda.cpp [NFC] polish colossalai/kernel/cuda_native/csrc/moe_cuda.cpp code style (#642) 3 years ago
moe_cuda_kernel.cu fix format (#583) 3 years ago
multi_tensor_adam.cu [NFC] polish colossalai/kernel/cuda_native/csrc/multi_tensor_adam.cu code style (#667) 3 years ago
multi_tensor_apply.cuh refactor kernel (#142) 3 years ago
multi_tensor_l2norm_kernel.cu [NFC] polish colossalai/kernel/cuda_native/csrc/multi_tensor_l2norm_kernel.cu code style (#635) 3 years ago
multi_tensor_lamb.cu [NFC] polish colossalai/kernel/cuda_native/csrc/multi_tensor_lamb.cu code stype (#628) 3 years ago
multi_tensor_scale_kernel.cu fix format (#563) 3 years ago
multi_tensor_sgd_kernel.cu refactor kernel (#142) 3 years ago
multihead_attention_1d.cpp add colossalai kernel module (#55) 3 years ago
multihead_attention_1d.h add colossalai kernel module (#55) 3 years ago
scaled_masked_softmax.cpp add colossalai kernel module (#55) 3 years ago
scaled_masked_softmax.h add colossalai kernel module (#55) 3 years ago
scaled_masked_softmax_cuda.cu add colossalai kernel module (#55) 3 years ago
scaled_upper_triang_masked_softmax.cpp add colossalai kernel module (#55) 3 years ago
scaled_upper_triang_masked_softmax.h add colossalai kernel module (#55) 3 years ago
scaled_upper_triang_masked_softmax_cuda.cu add colossalai kernel module (#55) 3 years ago
type_shim.h [cuda] modify the fused adam, support hybrid of fp16 and fp32 (#497) 3 years ago