xlm-roberta-base-finetuned-marc

This model is a fine-tuned version of xlm-roberta-base on the dutch_social dataset. It achieves the following results on the evaluation set:

• Loss: 0.1992
• Mae: 0.0532

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: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss	Mae
0.2824	1.0	10176	0.2370	0.0748
0.1809	2.0	20352	0.1992	0.0532

Framework versions

• Transformers 4.25.1
• Pytorch 1.13.0+cu116
• Datasets 2.8.0
• Tokenizers 0.13.2