File size: 1,216 Bytes
c245001
 
9c8f46b
 
136c794
735d4d9
 
586386b
d5f7662
 
 
7c50e44
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
586386b
d5f7662
 
586386b
 
 
 
 
d5f7662
 
 
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
33
34
35
36
37
38
39
---
license: apache-2.0
datasets:
- ehartford/wizard_vicuna_70k_unfiltered
base_model: OpenLLaMA-7B
---

# Overview
Fine-tuned [OpenLLaMA-7B](https://huggingface.co/openlm-research/open_llama_7b) with an uncensored/unfiltered Wizard-Vicuna conversation dataset [ehartford/wizard_vicuna_70k_unfiltered](https://huggingface.co/datasets/ehartford/wizard_vicuna_70k_unfiltered).
Used QLoRA for fine-tuning. Trained for one epoch on a 24GB GPU (NVIDIA A10G) instance, took ~18 hours to train.

# Prompt style
The model was trained with the following prompt style:
```
### HUMAN:
Hello

### RESPONSE:
Hi, how are you?

### HUMAN:
I'm fine.

### RESPONSE:
How can I help you?
...
```

# Training code
Code used to train the model is available [here](https://github.com/georgesung/llm_qlora).

# Demo
For a Gradio chat application using this model, clone [this HuggingFace Space](https://huggingface.co/spaces/georgesung/open_llama_7b_qlora_uncensored_chat/tree/main) and run it on top of a GPU instance.
The basic T4 GPU instance will work.

# Blog post
Since this was my first time fine-tuning an LLM, I also wrote an accompanying blog post about how I performed the training :)

https://georgesung.github.io/ai/qlora-ift/