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

move final progress_callback call to load_all_data

Browse source
This commit is contained in:
slaren 2023-12-20 23:23:21 +01:00
parent ecb23d4ac5
commit 8ed2a8eb62
1 changed files with 4 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

8
llama.cpp
View file

@ -2435,6 +2435,10 @@ struct llama_model_loader {
mapping->unmap(0, mmap_first); mapping->unmap(0, mmap_first);
mapping->unmap(mmap_last, mapping->size - mmap_last); mapping->unmap(mmap_last, mapping->size - mmap_last);
} }
if (progress_callback) {
progress_callback(1.0f, progress_callback_user_data);
}
} }
}; };
@ -3691,10 +3695,6 @@ static void llm_load_tensors(
ml.load_all_data(ctx, progress_callback, progress_callback_user_data, buf_mmap, use_mlock ? &model.mlock_mmap : NULL); ml.load_all_data(ctx, progress_callback, progress_callback_user_data, buf_mmap, use_mlock ? &model.mlock_mmap : NULL);
if (progress_callback) {
progress_callback(1.0f, progress_callback_user_data);
}
model.mapping = std::move(ml.mapping); model.mapping = std::move(ml.mapping);
// loading time will be recalculate after the first eval, so // loading time will be recalculate after the first eval, so