roberta-base-clickbait-post

This model is a fine-tuned version of FacebookAI/roberta-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.4912
• Accuracy: 0.735

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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	200	0.9791	0.68
No log	2.0	400	0.9764	0.7225
0.3769	3.0	600	1.0072	0.73
0.3769	4.0	800	1.2047	0.7075
0.2506	5.0	1000	1.4719	0.7
0.2506	6.0	1200	1.5090	0.72
0.2506	7.0	1400	1.7088	0.745
0.1383	8.0	1600	1.8566	0.7125
0.1383	9.0	1800	2.2784	0.7125
0.0756	10.0	2000	2.3264	0.705
0.0756	11.0	2200	2.2328	0.72
0.0756	12.0	2400	2.2335	0.73
0.0513	13.0	2600	2.3440	0.725
0.0513	14.0	2800	2.4086	0.725
0.0268	15.0	3000	2.3981	0.7275
0.0268	16.0	3200	2.4274	0.73
0.0268	17.0	3400	2.4699	0.735
0.0094	18.0	3600	2.4737	0.7375
0.0094	19.0	3800	2.4979	0.7225
0.0112	20.0	4000	2.4912	0.735

Framework versions

• Transformers 4.44.0.dev0
• Pytorch 2.4.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1