pansophic

/

rocket-3B

like 82

This change is intended for anyone trying to use llama.cpp to merge and quantize loras of this model. Thanks for creating such an awesome small model!

• Add optional parameter "rope_pct" (defaults to 1) to be compatible with https://github.com/ggerganov/llama.cpp/blob/master/convert-hf-to-gguf.py to allow for the conversion from safetensors to gguf.