Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
llama: fix memory leak for grammar (#10696)
Parth Sareen8cc33f4c2ba9
diff --git a/llama/sampling_ext.cpp b/llama/sampling_ext.cppindex 78b889bd7..1cacd3015 100644--- a/llama/sampling_ext.cpp+++ b/llama/sampling_ext.cpp@@ -114,6 +114,9 @@ void grammar_free(struct llama_grammar *g) {if (g->vocab != nullptr) {delete g->vocab;}+ if (g->o_vocab != nullptr) {+ delete g->o_vocab;+ }llama_grammar_free_impl(g);}}