tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-07-17 16:19:46 +00:00
Code Issues Packages Projects Releases Wiki Activity

graph : add llama_graph_result

Browse Source 
ggml-ci
This commit is contained in:
Georgi Gerganov
2025-02-18 11:16:53 +02:00
parent f0d3ff2388
commit c23590319a
5 changed files with 167 additions and 350 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
src/llama-context.h
View File

@ -95,6 +95,9 @@ struct llama_context : public llama_graph_i {
// zero-out inputs and create ggml_context
virtual ggml_context_ptr graph_init();
// TODO: add encode/decode graphs
virtual llama_graph_result graph_build(const llama_ubatch & ubatch, bool worst_case);
// returns the result of ggml_backend_sched_graph_compute_async execution
virtual enum ggml_status graph_compute(
ggml_cgraph * graph,
@ -145,9 +148,6 @@ struct llama_context : public llama_graph_i {
const llama_ubatch & ubatch,
int il);
// TODO: add encode/decode graphs
virtual ggml_cgraph * build_graph(const llama_ubatch & ubatch, bool worst_case);
// apply control vector for layer il
virtual ggml_tensor * build_cvec(
ggml_context * ctx0,