Public/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-09-17 08:19:38 +02:00
Code Issues Packages Projects Releases Wiki Activity

spec : add DFlash support (#22105)

Browse Source 
* spec: add DFlash v2 support

* dflash: support sliding window attention per layer_types

* docs: add dflash section

---------

Co-authored-by: Kashif Rasul <kashif.rasul@gmail.com>
This commit is contained in:
Ruixiang Wang
2026-06-28 15:01:34 +02:00
committed by GitHub
parent c1a1c8ee94
commit d1b34251bc
14 changed files with 712 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/llama-context.cpp
+2 -2
View File
@@ -100,10 +100,10 @@ llama_context::llama_context(
cparams.ctx_other = params.ctx_other;
}
if (model.arch == LLM_ARCH_EAGLE3) {
if (model.arch == LLM_ARCH_EAGLE3 || model.arch == LLM_ARCH_DFLASH) {
if (model.tok_embd == nullptr || model.output == nullptr) {
if (params.ctx_other == nullptr) {
throw std::runtime_error("EAGLE3 requires ctx_other to be set (this warning is normal during memory fitting)");
throw std::runtime_error(model.arch_name() + " requires ctx_other to be set (this warning is normal during memory fitting)");
}
cparams.ctx_other = params.ctx_other;
}