https://github.com/neoheartbeats/sthenno-chatbot

Frontend used to provide a Human-AI interaction environment
https://github.com/neoheartbeats/sthenno-chatbot

chatbot fine-tuning llm

Last synced: about 1 year ago
JSON representation

Frontend used to provide a Human-AI interaction environment

• Host: GitHub
• URL: https://github.com/neoheartbeats/sthenno-chatbot
• Owner: neoheartbeats
• Created: 2024-08-15T13:15:37.000Z (almost 2 years ago)
• Default Branch: main
• Last Pushed: 2024-10-24T10:00:11.000Z (over 1 year ago)
• Last Synced: 2025-04-10T20:18:13.076Z (about 1 year ago)
• Topics: chatbot, fine-tuning, llm
• Language: Python
• Homepage:
• Size: 438 KB
• Stars: 2
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# sthenno-chatbot

[![Code style: black](https://img.shields.io/badge/code%20style-black-000000.svg)](https://github.com/psf/black)

This framework is just currently a frontend used to provide a Human-AI interaction

environment. Endpoints are hosted using [vLLM](https://github.com/vllm-project/vllm).

For more details, please check [neoheartbeats-kernel](https://github.com/neoheartbeats/neoheartbeats-kernel).

## Current hosted models:

sthenno-gm-04 is a fine-tuned version of DeepMind's gemma2-9b-it.

This model is optimized by KTO(Kahneman-Tversky Optimization) using custom data.

This model is designed to output more naturally that to align human's preferences, but

NOT including to instruct the model to generate human-like outputs such as emotions.

One part of this design is to discover how LLMs implement mental models for

continual-learning and long-term memory's constructions.

Model's safetensors and training data have NOT been disclosed yet but planned to be by

publishing to platforms such as HuggingFace once reliable data is collected under

replicated evaluations.