Merge pull request #5310 from hpcaitech/feature/npu

Feature/npu

Merge pull request #5310 from hpcaitech/feature/npu
Feature/npu
8823cc48 · Frank Lee · GitHub · bce9499e · 73f4dc57 · 8823cc48
Unverified Commit 8823cc48 authored Jan 29, 2024 by Frank Lee Committed by GitHub Jan 29, 2024
20 changed files
--- a/extensions/__init__.py
+++ b/extensions/__init__.py
--- a/extensions/base_extension.py
+++ b/extensions/base_extension.py
--- a/extensions/cpp_extension.py
+++ b/extensions/cpp_extension.py
--- a/extensions/cpu_adam/__init__.py
+++ b/extensions/cpu_adam/__init__.py
--- a/op_builder/arm_cpu_adam.py
+++ b/op_builder/arm_cpu_adam.py
--- a/op_builder/cpu_adam.py
+++ b/op_builder/cpu_adam.py
--- a/colossalai/kernel/cuda_native/__init__.py
+++ b/colossalai/kernel/cuda_native/__init__.py
--- a/colossalai/kernel/cuda_native/csrc/cpu_adam_arm.cpp
+++ b/colossalai/kernel/cuda_native/csrc/cpu_adam_arm.cpp
--- a/colossalai/kernel/cuda_native/csrc/cpu_adam_arm.h
+++ b/colossalai/kernel/cuda_native/csrc/cpu_adam_arm.h
--- a/colossalai/kernel/cuda_native/csrc/colossal_C_frontend.cpp
+++ b/colossalai/kernel/cuda_native/csrc/colossal_C_frontend.cpp
--- a/colossalai/kernel/cuda_native/csrc/compat.h
+++ b/colossalai/kernel/cuda_native/csrc/compat.h
--- a/colossalai/kernel/cuda_native/csrc/cpu_adam.cpp
+++ b/colossalai/kernel/cuda_native/csrc/cpu_adam.cpp
--- a/colossalai/kernel/cuda_native/csrc/cpu_adam.h
+++ b/colossalai/kernel/cuda_native/csrc/cpu_adam.h
--- a/colossalai/kernel/cuda_native/csrc/kernels/include/block_reduce.h
+++ b/colossalai/kernel/cuda_native/csrc/kernels/include/block_reduce.h
--- a/colossalai/kernel/cuda_native/csrc/layer_norm_cuda.cpp
+++ b/colossalai/kernel/cuda_native/csrc/layer_norm_cuda.cpp
--- a/colossalai/kernel/cuda_native/csrc/layer_norm_cuda_kernel.cu
+++ b/colossalai/kernel/cuda_native/csrc/layer_norm_cuda_kernel.cu
--- a/colossalai/kernel/cuda_native/csrc/moe_cuda.cpp
+++ b/colossalai/kernel/cuda_native/csrc/moe_cuda.cpp
--- a/colossalai/kernel/cuda_native/csrc/moe_cuda_kernel.cu
+++ b/colossalai/kernel/cuda_native/csrc/moe_cuda_kernel.cu
--- a/colossalai/kernel/cuda_native/csrc/multi_tensor_adam.cu
+++ b/colossalai/kernel/cuda_native/csrc/multi_tensor_adam.cu
--- a/colossalai/kernel/cuda_native/csrc/multi_tensor_apply.cuh
+++ b/colossalai/kernel/cuda_native/csrc/multi_tensor_apply.cuh