happyz
/
llama.cpp
mirror of https://github.com/ggerganov/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Change due to ggml cgraph changes, all device work

This commit is contained in:
Yu, Zijun 2025-06-16 13:20:11 +08:00 committed by Mustafa Cavus
parent e27738a987
commit 42d4240937
1 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
ggml/src/ggml-openvino/ggml-decoder.cpp
View File

@ -216,9 +216,9 @@ void GgmlOvDecoder::set_input_output(ggml_tensor* node) {
void GgmlOvDecoder::set_max_token_len() {
for (int i = 0; i < m_cgraph->n_nodes; i++) {
auto* node = m_cgraph->nodes[i];
if (std::string(node->name) == "k-0") {
if (std::string(node->name) == "cache_k_l0 (view)") {
auto* cache_k = node->src[0];
m_max_token_len = cache_k->ne[0] / node->ne[0] / node->ne[2];
m_max_token_len = cache_k->ne[1];
break;
}
}