replica
/
llama.cpp
mirror of https://github.com/ggml-org/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

graph : fix graph reuse reset of params (#14760) 

ggml-ci
This commit is contained in:
Georgi Gerganov 2025-07-18 20:08:33 +03:00 committed by GitHub
parent 2adf8d83ac
commit 9fb1042ce6
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 2 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
src/llama-graph.cpp
View File

@ -428,6 +428,8 @@ void llm_graph_result::reset() {
t_embd = nullptr; t_embd = nullptr;
t_embd_pooled = nullptr; t_embd_pooled = nullptr;
params = {};
inputs.clear(); inputs.clear();
buf_compute_meta.resize(ggml_tensor_overhead()*max_nodes + ggml_graph_overhead_custom(max_nodes, false)); buf_compute_meta.resize(ggml_tensor_overhead()*max_nodes + ggml_graph_overhead_custom(max_nodes, false));