vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

CUDA GPU acceleration for LoRAs + f16 models (#1970)

Browse source
This commit is contained in:
Johannes Gäßler 2023-06-28 18:35:54 +02:00 • committed by GitHub
parent cfa0750bc9
commit 7f9753fa12
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
4 changed files with 78 additions and 19 deletions
Download patch file Download diff file Expand all files Collapse all files

7
examples/common.cpp
View file

@ -416,13 +416,6 @@ bool gpt_params_parse(int argc, char ** argv, gpt_params & params) {
exit(1);
}
#ifdef GGML_USE_CUBLAS
if (!params.lora_adapter.empty() && params.n_gpu_layers > 0) {
fprintf(stderr, "%s: error: the simultaneous use of LoRAs and GPU acceleration is not supported", __func__);
exit(1);
}
#endif // GGML_USE_CUBLAS
if (escape_prompt) {
process_escapes(params.prompt);
}