vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2025-09-10 09:04:36 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

fixed wrong sized struct from legacy q8_1, fixed opencl varsize arrays

Browse source
This commit is contained in:
Concedo 2023-05-13 23:56:08 +08:00
parent c9eb2ba1c5
commit e05455f852
7 changed files with 25 additions and 16 deletions
Download patch file Download diff file Expand all files Collapse all files

2
gpttype_adapter.cpp
View file

@ -224,7 +224,7 @@ ModelLoadResult gpttype_load_model(const load_model_inputs inputs, FileFormat in
neox_ctx.hparams.n_ctx = gptj_ctx_v1.hparams.n_ctx = gptj_ctx_v2.hparams.n_ctx = gpt2_ctx_v1.hparams.n_ctx = gpt2_ctx_v2.hparams.n_ctx = params.n_ctx;
printf("System Info: %s\n", llama_print_system_info());
SetQuantsUnshuffled(false);
if(file_format == FileFormat::GGML || file_format == FileFormat::GGHF || file_format == FileFormat::GGJT || file_format == FileFormat::GGJT_2)
{
//newer format has bit unshuffling