Upload README.md with huggingface_hub
Browse files
README.md
ADDED
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
![halos](assets/thumbnail.jpg)
|
2 |
+
|
3 |
+
This repo contains the model checkpoints for:
|
4 |
+
- model family pythia12-0b
|
5 |
+
- optimized with the loss kto
|
6 |
+
- aligned using the SHP, Anthropic HH and Open Assistant datasets.
|
7 |
+
|
8 |
+
Please refer to our code repository which contains intructions for training your own HALOs and links to our model cards.
|
9 |
+
|
10 |
+
If you find this repo or the technical paper useful in your research, please feel free to cite [our work](http://halos.github.io/):
|
11 |
+
```
|
12 |
+
@misc{ethayarajh2023halos,
|
13 |
+
url = {http://halos.github.io/},
|
14 |
+
author = {Ethayarajh, Kawin and Xu, Winnie, and Jurafsky, Dan and Kiela, Douwe},
|
15 |
+
title = {Human-Centered Loss Functions (HALOs)},
|
16 |
+
publisher = {Contextual AI Blog},
|
17 |
+
year = {2023},
|
18 |
+
}
|
19 |
+
```
|