nikola/llama_cpp
1
0
Fork 0
Code Issues Pull Requests Actions 10 Packages Projects Releases Wiki Activity

llama : deprecate llama_set_warmup (#24009)

Browse Source 
* llama : deprecate `llama_set_warmup`

* cont : fix type

Co-authored-by: Daniel Bevenius <daniel.bevenius@gmail.com>

---------

Co-authored-by: Daniel Bevenius <daniel.bevenius@gmail.com>
This commit is contained in:
Georgi Gerganov
2026-06-02 10:30:38 +03:00
committed by GitHub
co-authored by Daniel Bevenius
parent 8f7f3bf141
commit 4f3a4beb8d
4 changed files with 6 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files
include/llama.h
+5 -1
View File
@@ -976,7 +976,11 @@ extern "C" {
// Set whether the model is in warmup mode or not
// If true, all model tensors are activated during llama_decode() to load and cache their weights.
LLAMA_API void llama_set_warmup(struct llama_context * ctx, bool warmup);
//
// note: using this can cause extra graph reallocations because it changes the graph topology with MoE models,
// so it is generally not recommended to use in practice. will be removed in the future
DEPRECATED(LLAMA_API void llama_set_warmup(struct llama_context * ctx, bool warmup),
"user code should do warmup runs manually [TAG_LLAMA_GRAPH_NO_WARMUP]");
// Set abort callback
LLAMA_API void llama_set_abort_callback(struct llama_context * ctx, ggml_abort_callback abort_callback, void * abort_callback_data);