CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-07 09:57:00 +00:00
Code Issues Packages Projects Releases Wiki Activity

llama: print memory breakdown on exit (#15860)

Browse Source 
* llama: print memory breakdown on exit
This commit is contained in:
Johannes Gäßler
2025-09-24 16:53:48 +02:00
committed by GitHub
parent f2a789e334
commit e789095502
18 changed files with 243 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
src/llama-context.h
View File

@@ -17,9 +17,17 @@ class llama_batch_allocr;
class llama_io_read_i;
class llama_io_write_i;
// "memory" as in abstract memory for the context
struct llama_memory_i;
struct llama_memory_context_i;
// "memory" as in physical memory for a buffer type, in bytes
struct llama_memory_breakdown_data {
size_t model = 0; // memory allocated for the model
size_t context = 0; // memory allocated for the context
size_t compute = 0; // memory allocated for temporary compute buffers
};
struct llama_context {
// init scheduler and compute buffers, reserve worst-case graphs
llama_context(
@@ -144,6 +152,8 @@ struct llama_context {
llama_perf_context_data perf_get_data() const;
void perf_reset();
std::map<ggml_backend_buffer_type_t, llama_memory_breakdown_data> memory_breakdown() const;
//
// training
//