Files

 main

/

csrc

/

Copy path

Directory actions

More options

Directory actions

More options

Latest commit

mawong-amd

[Hotfix][CI/Build][Kernel] CUDA 11.8 does not support layernorm optim…

Apr 8, 2024

59a6abf · Apr 8, 2024

History

History

This branch is 4455 commits behind vllm-project/vllm:main.

/

csrc

/

Top

Folders and files

Name		Name	Last commit message	Last commit date
parent directory ..
attention		attention	Enable scaled FP8 (e4m3fn) KV cache on ROCm (AMD GPU) (vllm-project#3290	Apr 3, 2024
cpu		cpu	[Bugfix] Add kv_scale input parameter to CPU backend (vllm-project#3840)	Apr 4, 2024
moe		moe	Add fused top-K softmax kernel for MoE (vllm-project#2769)	Feb 6, 2024
punica		punica	[Kernel] support non-zero cuda devices in punica kernels (vllm-projec…	Mar 27, 2024
quantization		quantization	Enable scaled FP8 (e4m3fn) KV cache on ROCm (AMD GPU) (vllm-project#3290	Apr 3, 2024
activation_kernels.cu		activation_kernels.cu	Add kernel for GeGLU with approximate GELU (vllm-project#3337)	Mar 13, 2024
cache.h		cache.h	Enable scaled FP8 (e4m3fn) KV cache on ROCm (AMD GPU) (vllm-project#3290	Apr 3, 2024
cache_kernels.cu		cache_kernels.cu	Enable scaled FP8 (e4m3fn) KV cache on ROCm (AMD GPU) (vllm-project#3290	Apr 3, 2024
cuda_compat.h		cuda_compat.h	[ROCM] Fix blockReduceSum to use correct warp counts for ROCm and CUDA (	Mar 10, 2024
cuda_utils.h		cuda_utils.h	[ROCm] add support to ROCm 6.0 and MI300 (vllm-project#2274)	Jan 26, 2024
cuda_utils_kernels.cu		cuda_utils_kernels.cu	[ROCm] add support to ROCm 6.0 and MI300 (vllm-project#2274)	Jan 26, 2024
custom_all_reduce.cu		custom_all_reduce.cu	[BugFix] Some fixes for custom allreduce kernels (vllm-project#2760)	Mar 22, 2024
custom_all_reduce.cuh		custom_all_reduce.cuh	[BugFix] Some fixes for custom allreduce kernels (vllm-project#2760)	Mar 22, 2024
custom_all_reduce_test.cu		custom_all_reduce_test.cu	[BugFix] Some fixes for custom allreduce kernels (vllm-project#2760)	Mar 22, 2024
dispatch_utils.h		dispatch_utils.h	DeepseekMoE support with Fused MoE kernel (vllm-project#2453)	Jan 30, 2024
layernorm_kernels.cu		layernorm_kernels.cu	[Hotfix][CI/Build][Kernel] CUDA 11.8 does not support layernorm optim…	Apr 8, 2024
moe_align_block_size_kernels.cu		moe_align_block_size_kernels.cu	[Bugfix] Make moe_align_block_size AMD-compatible (vllm-project#3470)	Mar 18, 2024
ops.h		ops.h	Enable scaled FP8 (e4m3fn) KV cache on ROCm (AMD GPU) (vllm-project#3290	Apr 3, 2024
pos_encoding_kernels.cu		pos_encoding_kernels.cu	Add batched RoPE kernel (vllm-project#3095)	Mar 13, 2024
pybind.cpp		pybind.cpp	Enable scaled FP8 (e4m3fn) KV cache on ROCm (AMD GPU) (vllm-project#3290	Apr 3, 2024
reduction_utils.cuh		reduction_utils.cuh	[Kernel] Layernorm performance optimization (vllm-project#3662)	Mar 30, 2024
View all files