BERTje_finetuned_1986_1994

This model is a fine-tuned version of GroNLP/bert-base-dutch-cased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0941
• F1: 0.2000
• Precision: 0.1601
• Recall: 0.2666

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 2e-05
• train_batch_size: 3
• eval_batch_size: 3
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 6
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Precision	Recall
No log	1.0	203	0.1362	0.0907	0.0492	0.5748
No log	2.0	407	0.1035	0.1958	0.1463	0.2957
0.1888	3.0	610	0.0968	0.2000	0.1601	0.2666
0.1888	4.0	814	0.0946	0.2000	0.1601	0.2666
0.0974	4.99	1015	0.0941	0.2000	0.1601	0.2666

Framework versions

• Transformers 4.35.0
• Pytorch 1.11.0+cu102
• Datasets 2.14.6
• Tokenizers 0.14.1