![]() * llama : add thread safety test * llamafile : remove global state * llama : better LLAMA_SPLIT_MODE_NONE logic when main_gpu < 0 GPU devices are not used --------- Co-authored-by: Georgi Gerganov <ggerganov@gmail.com> |
||
---|---|---|
.. | ||
actions | ||
ISSUE_TEMPLATE | ||
workflows | ||
labeler.yml | ||
pull_request_template.md |