nikola/llama_cpp
1
0
Fork 0
Code Issues Pull Requests Actions 18 Packages Projects Releases Wiki Activity

fit : wrap llama_device_memory_data (#24522)

Browse Source
This commit is contained in:
Georgi Gerganov
2026-06-13 08:09:52 +03:00
committed by GitHub
parent c34b92235b
commit d8a24ccee2
3 changed files with 40 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files
common/fit.h
+10 -4
View File
@@ -34,12 +34,18 @@ void common_fit_print(
void common_memory_breakdown_print(const llama_context * ctx);
// TODO: convert this to common_device_memory_data that wraps llama_device_memory_data
// add API for accessing the internal `llama-ext.h` information
struct llama_device_memory_data;
struct common_device_memory_data {
int64_t total;
int64_t free;
size_t model;
size_t context;
size_t compute;
};
using common_device_memory_data_vec = std::vector<common_device_memory_data>;
// Load a model + context with no_alloc and return the per-device memory breakdown.
std::vector<llama_device_memory_data> common_get_device_memory_data(
common_device_memory_data_vec common_get_device_memory_data(
const char * path_model,
const llama_model_params * mparams,
const llama_context_params * cparams,