Hugging Face's logo

PJMixers
/
LLaMa-3-Instruct-ToxicQAFinal-ORPO-8B-QDoRA

Safetensors
Not-For-All-Audiences
Model card Files Community
LLaMa-3-Instruct-ToxicQAFinal-ORPO-8B-QDoRA
File size: 886 Bytes 
a0d7ed3
 
6158b66
 
 
a0d7ed3
 
5f96be1
411ef3d
db63d36
411ef3d
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
 
---
datasets:
- >-
  PJMixers/NobodyExistsOnTheInternet_ToxicQAFinal-L3-Instruct-8B-PreferenceShareGPT
- NobodyExistsOnTheInternet/ToxicQAFinal
tags:
- not-for-all-audiences
---
Trained on [NobodyExistsOnTheInternet/ToxicQAFinal](https://huggingface.co/datasets/NobodyExistsOnTheInternet/ToxicQAFinal). I converted the set to a preference dataset using refusals generated from LLaMa-3-Instruct-8B.

![train/rewards](https://huggingface.co/PJMixers/LLaMa-3-Instruct-ToxicQAFinal-ORPO-8B-QDoRA/resolve/main/images/rewards.png)
![train/logits](https://huggingface.co/PJMixers/LLaMa-3-Instruct-ToxicQAFinal-ORPO-8B-QDoRA/resolve/main/images/logits.png)
![train/logps](https://huggingface.co/PJMixers/LLaMa-3-Instruct-ToxicQAFinal-ORPO-8B-QDoRA/resolve/main/images/logps.png)
![train](https://huggingface.co/PJMixers/LLaMa-3-Instruct-ToxicQAFinal-ORPO-8B-QDoRA/resolve/main/images/train.png)