thek0tyara/llama-cpp-turboquant
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 6

llama : add use_direct_io flag for model loading (#18166)

Browse source 
* Adding --direct-io flag for model loading

* Fixing read_raw() calls

* Fixing Windows read_raw_at

* Changing type off_t to size_t for windows and Renaming functions

* disable direct io when mmap is explicitly enabled

* Use read_raw_unsafe when upload_backend is available, not functional on some devices with Vulkan and SYCL

* Fallback to std::fread in case O_DIRECT fails due to bad address

* Windows: remove const keywords and unused functions

* Update src/llama-mmap.cpp

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>

---------

Co-authored-by: jtischbein <jtischbein@gmail.com>
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
This commit is contained in:
Julius Tischbein 2026-01-08 07:35:30 +01:00 • committed by GitHub
parent 568371a726
commit 2038101bd9
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
12 changed files with 118 additions and 53 deletions
Download patch file Download diff file Expand all files Collapse all files

2
src/llama-quant.cpp
View file

@ -596,7 +596,7 @@ static void llama_model_quantize_impl(const std::string & fname_inp, const std::
}
std::vector<std::string> splits = {};
llama_model_loader ml(fname_inp, splits, use_mmap, /*check_tensors*/ true, /*no_alloc*/ false, kv_overrides, nullptr);
llama_model_loader ml(fname_inp, splits, use_mmap, /*use_direct_io*/ true, /*check_tensors*/ true, /*no_alloc*/ false, kv_overrides, nullptr);
ml.init_mappings(false); // no prefetching
llama_model model(llama_model_default_params());