File size: 4,129 Bytes
71479e4
 
 
 
 
0dea749
 
71479e4
0dea749
 
71479e4
6401b5f
71479e4
0dea749
 
 
145ccea
71479e4
 
 
0dea749
6401b5f
0dea749
71479e4
3bb946e
6401b5f
3927ae9
0dea749
 
7c68121
24ab3b7
6401b5f
 
 
 
340479a
7c68121
6401b5f
7c68121
340479a
 
0f00565
340479a
dc966c6
71479e4
 
 
0f00565
71479e4
0f00565
71479e4
0f00565
71479e4
 
 
0f00565
 
22e97d3
 
71479e4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0dea749
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
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
---
license: mit
base_model: microsoft/phi-2
tags:
- trl
- fietje
- alignment-handbook
datasets:
- uonlp/CulturaX
- wikimedia/wikipedia
model-index:
- name: fietje-2
  results: []
language:
- nl
pipeline_tag: text-generation
inference: false
---


<p align="center" style="margin:0;padding:0">
  <img src="https://huggingface.co/BramVanroy/fietje-2/resolve/main/img/fietje-2b-banner-rounded.png" alt="Fietje banner" width="800" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
</p>

<div style="margin:auto; margin-top: 0; text-align:center">
  <h1 style="margin-bottom: 0">Fietje 2</h1>
  <em>An open and efficient LLM for Dutch</em>
</div>

<blockquote class="tip" style="padding: 1.5em; border: 0">
  <p align="center" style="text-align: center; margin: 0">
    <a rel="nofollow" href="https://huggingface.co/BramVanroy/fietje-2">👱‍♀️ Base version</a> (this one) -
    <a rel="nofollow" href="https://huggingface.co/BramVanroy/fietje-2-instruct">🤖 Instruct version</a> -
    <a rel="nofollow" href="https://huggingface.co/BramVanroy/fietje-2-chat">💬 Chat version</a> -
    <a rel="nofollow" href="https://huggingface.co/BramVanroy/fietje-2-GGUF">🚀 GGUF of base</a>
  </p>
  <p align="center" style="text-align: center; margin: 0">
    <a href="https://huggingface.co/spaces/BramVanroy/fietje-2"><strong>Chat with Fietje here!</strong></a>
  </p>
</blockquote>

Fietje is an adapated version of [microsoft/phi-2](https://huggingface.co/microsoft/phi-2), tailored to Dutch text generation by training on 28B tokens. It is small and efficient with a size of 2.7 billion parameters while performing almost on par with more powerful Dutch LLMs of twice its size like [GEITje 7B Ultra](https://huggingface.co/BramVanroy/GEITje-7B-ultra).

A thorough description of the creation and evaluation of Fietje as well as usage examples are available in [this Github repository](https://github.com/BramVanroy/fietje).

## Intended uses & limitations

The same limitations as [phi-2](https://huggingface.co/microsoft/phi-2#limitations-of-phi-2), and LLMs in general, apply here. LLMs hallucinate, make mistakes, and should not be trusted. Use at your own risk!

## Training data

Fietje was continue-pretrained on 28B Dutch tokens, which includes the full Dutch component of Wikipedia (accounting for around 15%), supplemented with Dutch tokens from CulturaX. A newer version of this dataset can be found [here](https://huggingface.co/datasets/BramVanroy/wikipedia_culturax_dutch), which also describes the filtering that took place to ensure high data quality.

## Training procedure

I am thankful to the [Flemish Supercomputer Center](https://www.vscentrum.be/) (VSC) for providing the computational power to accomplish this project. Accounting for waiting for jobs, training took around two weeks on four nodes of 4x A100 80GB each (16 total).

Training was done with the wonderful [alignment-handbook](https://github.com/huggingface/alignment-handbook), using DeepSpeed as a back-end. Exact training recipes and SLURM script are given in the [Github repository](https://github.com/BramVanroy/fietje).

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 9e-05
- train_batch_size: 40
- eval_batch_size: 40
- seed: 42
- distributed_type: multi-GPU
- num_devices: 16
- gradient_accumulation_steps: 3
- total_train_batch_size: 1920
- total_eval_batch_size: 640
- optimizer: Adam with betas=(0.9,0.98) and epsilon=1e-07
- lr_scheduler_type: linear
- num_epochs: 1.0

### Training results

| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 1.6334        | 0.13  | 900  | 1.5937          |
| 1.5469        | 0.26  | 1800 | 1.5051          |
| 1.4937        | 0.4   | 2700 | 1.4628          |
| 1.4633        | 0.53  | 3600 | 1.4375          |
| 1.4485        | 0.66  | 4500 | 1.4203          |
| 1.4374        | 0.79  | 5400 | 1.4085          |
| 1.4278        | 0.92  | 6300 | 1.4013          |


### Framework versions

- Transformers 4.39.1
- Pytorch 2.1.2+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2