README.md

metadata

language:
  - en
license: apache-2.0
tags:
  - text-generation-inference
  - transformers
  - unsloth
  - llama
  - gguf
base_model: teknium/OpenHermes-7B
datasets:
  - 922-SY/MoCha_DPO_03022024_test1

Uploaded model

• Developed by: 922CA
• License: apache-2.0
• Finetuned from model : teknium/OpenHermes-7B

This llama model was trained 2x faster with Unsloth and Huggingface's TRL library.

Notes:

• Doesn't perform as well as older models (gets information wrong, too out-of-character, etc. - DPO dataset could be improved), looking into it