There has been a llama model file hosted on [Hugging Face](https://huggingface.co/decapoda-research/llama-30b-hf/tree/main) It would be good if there is a convert script for this format as well, just like what has been done on [whisper.cpp](https://github.com/ggerganov/whisper.cpp/blob/09e90680072d8ecdf02eaf21c393218385d2c616/models/convert-pt-to-ggml.py)
Activity
hpnyaggerman commentedon Mar 14, 2023
Just download the original models here: meta-llama/llama#73
Merge pull request ggml-org#117 from TheTom/fix/issue-81-nix-spirv-dupe
Merge pull request ggml-org#117 from rcfa/pr/qsa-vulkan