llama.cpp fork for Ternary Bonsai 2 27B: MTP speculation, suffix cache reuse (SCR), q8 KV, two-slot fairness, fused DeltaNet kernels
4 folders · 40 files

apple
hip
jinja
snapdragon
bench-models.sh
build-info.sh
ccache-clear.sh
check-requirements.sh
compare-commits.sh
compare-llama-bench.py
compare-logprobs.py
create_ops_docs.py
debug-test.sh
gen-authors.sh
gen-unicode-data.py
get_chat_template.py
get-flags.mk
get-hellaswag.sh
get-pg.sh
get-wikitext-2.sh
get-winogrande.sh
git-bisect-run.sh
git-bisect.sh
hf.sh
install-oneapi.bat
make-release-checks.sh
make-release-desc.sh
pr2wt.sh
release.sh
serve-static.js
server-bench.py
server-test-function-call.py
server-test-model.py
server-test-parallel-tc.py
server-test-structured.py
sync_vendor.py
sync-ggml-am.sh
sync-ggml.last
sync-ggml.sh
tool_bench.py
tool_bench.sh
ui-assets.cmake
verify-checksum-models.py
wc2wt.sh