eachadea's picture
Update README.md
1beaa85
|
raw
history blame
587 Bytes
metadata
pipeline_tag: conversational
tags:
  - vicuna
  - llama
  - text-generation-inference

Converted for use with llama.cpp

  • Based on AlekseyKorshuk/vicuna-7b
  • Uncensored
  • 4-bit quantized
  • Needs ~6GB of CPU RAM
  • Won't work with alpaca.cpp or old llama.cpp (new ggml format requires latest llama.cpp)
  • 7B parameter version
  • EOS token fix on revision 1

Bigger 13B version can be found here: https://huggingface.co/eachadea/ggml-vicuna-13b-4bit.

Unlike 7B, 13B is censored.


tags: - vicuna - llama - text-generation-inference