distilroberta-base-clickbait-task1-20-epoch-post

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

• Loss: 2.5503
• Accuracy: 0.6825

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	0.7686	0.69
No log	2.0	400	0.7584	0.705
0.7788	3.0	600	0.7580	0.6925
0.7788	4.0	800	0.8267	0.71
0.4168	5.0	1000	0.9697	0.7
0.4168	6.0	1200	1.1594	0.685
0.4168	7.0	1400	1.4171	0.6925
0.1774	8.0	1600	1.6601	0.715
0.1774	9.0	1800	1.8113	0.6825
0.0813	10.0	2000	1.9062	0.705
0.0813	11.0	2200	2.0308	0.7075
0.0813	12.0	2400	2.4102	0.6625
0.0419	13.0	2600	2.3119	0.695
0.0419	14.0	2800	2.3785	0.6825
0.0212	15.0	3000	2.4639	0.69
0.0212	16.0	3200	2.4754	0.69
0.0212	17.0	3400	2.5110	0.69
0.0166	18.0	3600	2.6399	0.6875
0.0166	19.0	3800	2.5551	0.6825
0.0108	20.0	4000	2.5503	0.6825

Framework versions

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