GEITje-7B-chat
[!CAUTION] ⚠️ At the pressing request of Stichting BREIN, GEITje is no longer available, starting immediately. ⚠️
All model files (the weights) and checkpoints have been deleted from this repo. See my blog post (Dutch, English) for further clarification.
🐐 Check out GEITje-7b-chat-v2 🐐
GEITje-7B
GEITje is a large open Dutch language model with 7 billion parameters, based on Mistral 7B. It has been further trained on 10 billion tokens of Dutch text. This has improved its Dutch language skills and increased its knowledge of Dutch topics.
Model description
Mistral – Base Model
GEITje is based on Mistral 7B. It's a large open language model with 7 billion parameters, trained by Mistral AI. According to Mistral AI, the 7B model performs better than Llama 2 13B on all (English-language) benchmarks they tested it on. Mistral 7B has been released under the Apache 2.0 open source license.
GEITje – Trained Further on Dutch Texts
GEITje was created by further training Mistral 7B on no less than 10 billion tokens of Dutch text from the Dutch Gigacorpus and the MADLAD-400 web crawling corpus. It is a so-called full-parameter finetune: performed on all parameters. It is not a PEFT or LoRA finetune. Like Mistral, GEITje has a context length of 8,192 tokens.
GEITje-chat – Finetuned for Dialogues
As a demonstration of GEITje's capabilities for chat applications, two initial chat variants of GEITje have also been finetuned: GEITje-chat and GEITje-chat-v2. They can follow instructions, answer questions, and hold dialogues on a variety of topics.
More info
Read more about GEITje-chat in the 📄 README on GitHub.
Checkpoints
Intermediate checkpoints are available in the checkpoints branch.
Training procedure
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: 3
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.0263 | 0.2 | 236 | 0.9482 |
| 1.0368 | 0.4 | 472 | 0.9574 |
| 0.9503 | 0.6 | 708 | 0.9492 |
| 1.1419 | 0.8 | 944 | 0.9406 |
| 1.2161 | 1.0 | 1180 | 0.9317 |
| 0.6695 | 1.2 | 1416 | 0.9407 |
| 0.7379 | 1.4 | 1652 | 0.9350 |
| 0.7695 | 1.6 | 1888 | 0.9282 |
| 0.6795 | 1.8 | 2124 | 0.9218 |
| 0.6217 | 2.0 | 2360 | 0.9174 |
| 0.438 | 2.2 | 2596 | 0.9546 |
| 0.3719 | 2.39 | 2832 | 0.9546 |
| 0.4853 | 2.59 | 3068 | 0.9548 |
| 0.3852 | 2.79 | 3304 | 0.9548 |
| 0.48 | 2.99 | 3540 | 0.9548 |
Framework versions
- Transformers 4.36.0.dev0
- Pytorch 2.1.1+cu121
- Datasets 2.15.0
- Tokenizers 0.15.0