Public/koboldcpp
1
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2026-09-01 10:00:51 +02:00
Code Issues Packages Projects Releases Wiki Activity

Merge commit '8c146a8366304c871efc26057cc90370ccf58dad' into concedo_experimental

Browse Source 
# Conflicts:
#	src/CMakeLists.txt
#	tests/test-llama-archs.cpp
This commit is contained in:
Concedo
2026-06-30 22:00:03 +08:00
parent 748a313997 8c146a8366
commit 61ad97cbc1
26 changed files with 4591 additions and 40 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/llama-context.cpp
+5 -1
View File
@@ -2331,7 +2331,11 @@ void llama_context::output_reorder() {
//
uint32_t llama_context::graph_max_nodes(uint32_t n_tokens) const {
if (model.arch == LLM_ARCH_QWEN3NEXT || model.arch == LLM_ARCH_KIMI_LINEAR || model.arch == LLM_ARCH_QWEN35 || model.arch == LLM_ARCH_QWEN35MOE) {
if (model.arch == LLM_ARCH_QWEN3NEXT ||
model.arch == LLM_ARCH_KIMI_LINEAR ||
model.arch == LLM_ARCH_QWEN35 ||
model.arch == LLM_ARCH_QWEN35MOE ||
model.arch == LLM_ARCH_DEEPSEEK4) {
return std::max<uint32_t>(n_tokens * 40, 32u * model.n_tensors());
}
uint32_t res = std::max<uint32_t>(1024u, 8u*model.n_tensors());