BERTje_finetuned_1986

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: 0.1358
• F1: 0.2445
• Precision: 0.1567
• Recall: 0.5560

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	93	0.2750	0.0628	0.0324	0.9965
No log	2.0	186	0.1739	0.0960	0.0509	0.8466
No log	3.0	279	0.1476	0.1795	0.1043	0.6426
No log	4.0	372	0.1383	0.2334	0.1469	0.5679
No log	5.0	465	0.1358	0.2445	0.1567	0.5560

Framework versions

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