metadata
datasets:
- GroNLP/dutch-cola
language:
- nl
tags:
- generated from trainer
metrics:
- accuracy
Model Card for Model ID
This model is a fine-tuned version of yhavinga/gpt2-medium-dutch on GroNLP/dutch-cola. It achieves the following results on the evaluation set:
- Loss: 0.519
- Accuracy: 0.7613
Training Details
Training Hyperparameters
The following hyperparameters were used during training:
- learning rate: 4e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: AdamW with lr=4e-05, weight_decay=0.01, betas=(o.9, 0.999) and epsilon=1e-08
- num_epochs=3
- fp16=True
- gradient_acc_step=1
Evaluation
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.5648 | 1.0 | 1244 | 0.5695 | 0.7192 |
| 0.3399 | 2.0 | 2488 | 0.5190 | 0.7613 |
| 0.1779 | 3.0 | 3732 | 0.7269 | 0.7625 |