tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-07-26 03:03:25 -04:00
Code Issues Packages Projects Releases Wiki Activity

metal : parallel RoPE on Metal (#3024)

Browse Source 
* Parallel RoPE on metal

* PR suggestion

---------

Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
This commit is contained in:
Kawrakow
2023-09-07 15:45:01 +02:00
committed by GitHub
parent be6beeb8d7
commit be8c9c245b
2 changed files with 15 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
ggml-metal.m
View File

@ -1141,7 +1141,7 @@ void ggml_metal_graph_compute(
[encoder setBytes:&freq_base length:sizeof(float) atIndex:21];
[encoder setBytes:&freq_scale length:sizeof(float) atIndex:22];
[encoder dispatchThreadgroups:MTLSizeMake(ne01, ne02, ne03) threadsPerThreadgroup:MTLSizeMake(1, 1, 1)];
[encoder dispatchThreadgroups:MTLSizeMake(ne01, ne02, ne03) threadsPerThreadgroup:MTLSizeMake(32, 1, 1)];
} break;
case GGML_OP_DUP:
case GGML_OP_CPY: