File size: 1,506 Bytes
87499a5
 
 
2ab8898
ba98fb8
2ab8898
ba98fb8
22b9989
 
ba98fb8
 
 
 
 
 
 
 
 
 
 
22b9989
ba98fb8
 
 
 
 
 
 
 
 
 
 
 
 
 
2ab8898
ba98fb8
2ab8898
ba98fb8
 
 
 
 
 
 
 
 
2ab8898
 
 
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
40
41
42
43
44
45
46
47
48
49
50
---
license: cc-by-4.0
---
# **KoQuality-Polyglot-5.8b**

KoQuality-Polyglot-5.8b is an auto-regressive language model that conducts instruction tuning with KoQuality datasets on Polyglot-5.8b model. 

<img src=https://cdn-uploads.huggingface.co/production/uploads/650fecfd247f564485f8fbcf/U7fLI5xFm23kosszXtyLP.png style="max-width: 500px; width: 300%"/>




| Model | 0-shot | 1-shot | 2-shot | 5-shot | 10-shot
| --- | --- | --- | --- | --- | --- |
| koquality-polyglot-5.8b | 0.5472 | 0.5979 | 0.6260 | 0.6486 | 0.6535
| polyglot-ko-5.8b | 0.5587 | 0.5977 | 0.6138 | 0.6431 | 0.6457
| koalpcaca-polyglot-5.8b | 0.5085 | 0.5561 | 0.5768 | 0.6097 | 0.6059
| kullm-polyglot-5.8b | 0.5409 | 0.6072 | 0.5945 | 0.6345 | 0.6530


<img src=https://cdn-uploads.huggingface.co/production/uploads/650fecfd247f564485f8fbcf/7EKl1OAgKgPBFcSlGzBiW.png style="max-width: 800px; width: 400%"/>


## Training hyperparameters
- learning_rate: 5e-5
- train_batch_size: 4
- seed: 42
- distributed_type: multi-GPU (A100 80G)
- num_devices: 4
- gradient_accumulation_steps: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2.0



## Citation

```
@misc{2023koqaulity,
  title = {KoQuality: Curation of High-quality Instruction Data for Korean Language Models},
  author = {Na, Yohan and Kim, Dahye and Chae, Dong-Kyu},
  journal={Proceedings of the 35th Annual Conference on Human and Cognitive Language Technology (HCLT 2023)},
  pages={},
  year = {2023},
}
```


<br>