vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2026-08-18 21:05:36 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

no need snapshot after gen is complete.

Browse source
This commit is contained in:
Concedo 2026-03-15 12:34:48 +08:00
parent 33211e6edf
commit fcdf2f40d5
1 changed files with 6 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

12
gpttype_adapter.cpp
View file

@ -2591,7 +2591,7 @@ ModelLoadResult gpttype_load_model(const load_model_inputs inputs, FileFormat in
llama_model * llamamodel = llama_model_load_from_file(kcpp_data->model_filename.c_str(), model_params);
//prepare savestate slots
//prepare savestate slots
savestate_limit = inputs.smartcacheslots;
//if RNN model AND shifting and fastforward is on, enable smartcache
@ -2601,7 +2601,7 @@ ModelLoadResult gpttype_load_model(const load_model_inputs inputs, FileFormat in
{
printf("RNN or Hyrbid model with FF and shifting flags enabled - SmartCache will be enabled with extra slots. Disable CtxShift if you do not want this.\n",savestate_limit);
kcpp_data->smartcache = true;
savestate_limit += 3;
savestate_limit += 2;
}
}
savestates.resize(savestate_limit);
@ -5160,10 +5160,10 @@ generation_outputs gpttype_generate(const generation_inputs inputs)
}
//if running rnn model in smartcache mode, save progress after each gen
if(kcpp_data->smartcache && is_recurrent && file_format==FileFormat::GGUF_GENERIC && current_context_tokens.size() > 32)
{
smartcache_quick_snapshot();
}
// if(kcpp_data->smartcache && is_recurrent && file_format==FileFormat::GGUF_GENERIC && current_context_tokens.size() > 32)
// {
// smartcache_quick_snapshot();
// }
if(debugmode==1 && !is_quiet && file_format == FileFormat::GGUF_GENERIC)
{