File size: 1,102 Bytes
44657c5
 
 
 
 
d102071
44657c5
 
 
eb4e457
 
44657c5
eb4e457
 
44657c5
 
eb4e457
44657c5
eb4e457
 
 
 
44657c5
eb4e457
44657c5
eb4e457
44657c5
eb4e457
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
---
library_name: transformers
tags: []
---

# koalpaca-polyglot-PEFT-ex

<!-- Provide a quick summary of what the model is/does. -->

This model is the result of the PEFT study.  
The study was conducted to understand how to fine-tune the open-source LLM using QLoRA.  

This model is not made for the general performance.  
So, it is not recommended to use it in practice.


## Model Details

This model is trained to answer the question about a certain library.  
The model is finetuned from "KoAlpaca".  
It only trained 0.099% of parameters, using the QLoRA method.  
Even Though it trained a small number of parameters, the model gives an actual answer about the data that was trained.

## Model Description

<!-- Provide a longer summary of what this model is. -->

- **Developed by:** [Chae Minsu](https://github.com/chaeminsoo)
- **Model type:** Text Generation
- **Language(s) (NLP):** Korean
- **Finetuned from model:** [beomi/polyglot-ko-12.8b-safetensors](https://huggingface.co/beomi/polyglot-ko-12.8b-safetensors)
- **Training Data:** Snippet of Kyungpook National University Library FAQ