da-sentiment

This model is a fine-tuned version of Maltehb/danish-bert-botxo on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.4976
• Accuracy: 0.8377

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: 1e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	58	0.6168	0.7446
No log	2.0	116	0.5689	0.7554
0.4255	3.0	174	0.5542	0.7814
0.4255	4.0	232	0.5224	0.7944
0.2655	5.0	290	0.5172	0.8030
0.2655	6.0	348	0.4992	0.8182
0.1944	7.0	406	0.4852	0.8290
0.1944	8.0	464	0.4972	0.8312
0.1772	9.0	522	0.4905	0.8333
0.1772	10.0	580	0.4915	0.8290
0.1772	11.0	638	0.4936	0.8333
0.1654	12.0	696	0.4939	0.8333
0.1654	13.0	754	0.4944	0.8377
0.1605	14.0	812	0.4969	0.8355
0.1605	15.0	870	0.4976	0.8377

Framework versions

• Transformers 4.28.0
• Pytorch 2.0.0+cu118
• Datasets 2.12.0
• Tokenizers 0.13.3