roberta-base-clickbait-task1-20-epoch-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: 2.4441
• Accuracy: 0.7275

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	0.7323	0.7475
No log	2.0	400	0.7165	0.735
0.7415	3.0	600	0.7385	0.7325
0.7415	4.0	800	0.8518	0.7125
0.3635	5.0	1000	0.9134	0.735
0.3635	6.0	1200	1.1249	0.7325
0.3635	7.0	1400	1.4414	0.715
0.1572	8.0	1600	1.5655	0.7325
0.1572	9.0	1800	1.7617	0.7225
0.0732	10.0	2000	2.0701	0.71
0.0732	11.0	2200	2.2284	0.705
0.0732	12.0	2400	2.1715	0.6975
0.036	13.0	2600	2.2769	0.715
0.036	14.0	2800	2.3432	0.715
0.0162	15.0	3000	2.3586	0.7225
0.0162	16.0	3200	2.3757	0.73
0.0162	17.0	3400	2.4715	0.715
0.013	18.0	3600	2.4563	0.7275
0.013	19.0	3800	2.4388	0.7275
0.0098	20.0	4000	2.4441	0.7275

Framework versions

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