LLM inference in C/C++
C++ ggml
llama : add guard for K/V rotation input when buffer is unallocated (#25215)
liminfei-amda4107133a634
diff --git a/src/llama-graph.cpp b/src/llama-graph.cppindex 4c86e43c1f..dc41d66904 100644--- a/src/llama-graph.cpp+++ b/src/llama-graph.cpp@@ -494,11 +494,11 @@ void llm_graph_input_attn_kv::set_input(const llama_ubatch * ubatch) {mctx->set_input_kq_mask(self_kq_mask, ubatch, cparams.causal_attn);}- if (self_k_rot) {+ if (self_k_rot && self_k_rot->buffer) {mctx->set_input_k_rot(self_k_rot);}- if (self_v_rot) {+ if (self_v_rot && self_v_rot->buffer) {mctx->set_input_v_rot(self_v_rot);}}@@ -592,19 +592,19 @@ void llm_graph_input_attn_kv_iswa::set_input(const llama_ubatch * ubatch) {mctx->get_swa()->set_input_kq_mask(self_kq_mask_swa, ubatch, cparams.causal_attn);}- if (self_k_rot) {+ if (self_k_rot && self_k_rot->buffer) {mctx->get_base()->set_input_k_rot(self_k_rot);}- if (self_v_rot) {+ if (self_v_rot && self_v_rot->buffer) {mctx->get_base()->set_input_v_rot(self_v_rot);}- if (self_k_rot_swa) {+ if (self_k_rot_swa && self_k_rot_swa->buffer) {mctx->get_swa()->set_input_k_rot(self_k_rot_swa);}- if (self_v_rot_swa) {+ if (self_v_rot_swa && self_v_rot_swa->buffer) {mctx->get_swa()->set_input_v_rot(self_v_rot_swa);}}