biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Kernel] Full Tensor Parallelism for LoRA Layers (#3524)

Browse Source 
Co-authored-by: Antoni Baum <antoni.baum@protonmail.com>
This commit is contained in:
Austin Veselka
2024-04-27 02:03:48 -05:00
committed by GitHub
parent 18d23f642a
commit eefeb16464
19 changed files with 686 additions and 111 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
csrc/punica/bgmv/bgmv_fp16_fp32_fp16.cu
View File

@@ -2,3 +2,4 @@
#include "bgmv_impl.cuh"
FOR_BGMV_WIDE_NARROW(INST_BGMV_TWOSIDE, nv_half, float, nv_half)
FOR_INST_BGMV_WIDE_NARROW(INST_BGMV_ONESIDE, nv_half, float, nv_half)