tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-06-28 04:15:21 +00:00
Code Issues Packages Projects Releases Wiki Activity

llama : make tensor_split ptr instead of array (#2272)

Browse Source
This commit is contained in:
Georgi Gerganov
2023-07-21 13:10:51 +03:00
committed by GitHub
parent 54e3bc76fe
commit ae178ab46b
4 changed files with 8 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
examples/common.cpp
View File

@ -586,7 +586,7 @@ struct llama_context_params llama_context_params_from_gpt_params(const gpt_param
lparams.n_batch = params.n_batch;
lparams.n_gpu_layers = params.n_gpu_layers;
lparams.main_gpu = params.main_gpu;
memcpy(lparams.tensor_split, params.tensor_split, LLAMA_MAX_DEVICES*sizeof(float));
lparams.tensor_split = params.tensor_split;
lparams.low_vram = params.low_vram;
lparams.seed = params.seed;
lparams.f16_kv = params.memory_f16;