File size: 638 Bytes
02fb3a5 81a070f 02fb3a5 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 |
---
title: DPO Demo
emoji: 📚
colorFrom: blue
colorTo: blue
sdk: gradio
sdk_version: 6.0.2
app_file: app.py
pinned: false
short_description: Testing DPO for finetuning models
---
A test / demo application playground for DPO Preference Tuning on different LLM models.
Running on Huggingspace:
https://huggingface.co/spaces/CatoG/DPO_Demo
Allows for LLM model selection, preference tuning of LLM responses, model response tuning with LoRA and Direct Preference Optimization (DPO).
Tuned model / policies can be downloaded for further use.
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
|