README.md

metadata

license: cc-by-nc-4.0
datasets:
  - tatsu-lab/alpaca
language:
  - en
metrics:
  - vicuna benchmark
  - wikitext2
pipeline_tag: question-answering

Eluwa: A Conversational LoRA for Facebook's OPT 2.7b Architecture

Eluwa is a fine-tuned LoRA model based on Facebook's OPT 2.7b architecture and trained on the Stanford Alpaca dataset. Eluwa is designed to provide a more conversational and creative experience in question-answering mode compared to the default OPT model. The idea was that OPT was too curt (and frankly, a bit of an asshole) for a model of its size, and that we could finetune it like Alpaca did to Llama.

It worked! Based on very limited testing, it's about halfway to GPT 3.5. Response times are fast: on my GTX 1080ti + Ryzen 3600,it generates between 1.14 tokens/s and 3.77 tokens/s.