4 ms·
I'm not able to run 13B and from his wiki: > Currently, only LLaMA-7B is supported since I haven't figured out how to merge the tensors of the bigger models.
by bojangleslover 4y ago
I'm not able to run 13B and from his wiki:
> Currently, only LLaMA-7B is supported since I haven't figured out how to merge the tensors of the bigger models.
- minxomat 4y agoThis has been fixed almost 2 days ago now. It’s literally mentioned at the top of the repo.
- simonw 4y agoThis commit landed 7 hours ago (since I wrote my TIL): https://github.com/ggerganov/llama.cpp/commit/007a8f6f459c6eb56678fdee4c09219ddb85b640 https://github.com/ggerganov/llama.cpp/commit/007a8f6f459c6e...