File size: 3,010 Bytes
78c5209 d4d13e3 78c5209 d4d13e3 78c5209 d4d13e3 78c5209 d4d13e3 78c5209 d4d13e3 78c5209 d4d13e3 78c5209 d4d13e3 |
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 |
---
license: apache-2.0
base_model: Rijgersberg/GEITje-7B-chat-v2
tags:
- trl
- sft
- generated_from_trainer
model-index:
- name: AmsterdamDocClassificationGEITje200T2Epochs
results: []
datasets:
- FemkeBakker/AmsterdamBalancedFirst200Tokens
language:
- nl
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# AmsterdamDocClassificationGEITje200T2Epochs
As part of the Assessing Large Language Models for Document Classification project by the Municipality of Amsterdam, we fine-tune Mistral, Llama, and GEITje for document classification.
The fine-tuning is performed using the [AmsterdamBalancedFirst200Tokens](https://huggingface.co/datasets/FemkeBakker/AmsterdamBalancedFirst200Tokens) dataset, which consists of documents truncated to the first 200 tokens.
In our research, we evaluate the fine-tuning of these LLMs across one, two, and three epochs.
This model is a fine-tuned version of [Rijgersberg/GEITje-7B-chat-v2](https://huggingface.co/Rijgersberg/GEITje-7B-chat-v2) and has been fine-tuned for two epochs.
It achieves the following results on the evaluation set:
- Loss: 0.5796
## Training and evaluation data
- The training data consists of 9900 documents and their labels formatted into conversations.
- The evaluation data consists of 1100 documents and their labels formatted into conversations.
## Training procedure
See the [GitHub](https://github.com/Amsterdam-Internships/document-classification-using-large-language-models) for specifics about the training and the code.
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 1e-05
- train_batch_size: 2
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 2
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:------:|:----:|:---------------:|
| 0.7664 | 0.1988 | 123 | 0.6890 |
| 0.6617 | 0.3976 | 246 | 0.6347 |
| 0.3825 | 0.5964 | 369 | 0.6028 |
| 0.4427 | 0.7952 | 492 | 0.5913 |
| 0.6739 | 0.9939 | 615 | 0.5906 |
| 0.4407 | 1.1939 | 738 | 0.5918 |
| 0.6671 | 1.3927 | 861 | 0.5835 |
| 0.4845 | 1.5915 | 984 | 0.5802 |
| 0.4699 | 1.7903 | 1107 | 0.5796 |
| 0.5434 | 1.9891 | 1230 | 0.5796 |
Training time: it took in total 1 hour and 36 minutes to fine-tune the model for two epochs.
### Framework versions
- Transformers 4.41.1
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1
### Acknowledgements
This model was trained as part of [insert thesis info] in collaboration with Amsterdam Intelligence for the City of Amsterdam. |