distilroberta-base-clickbait-task1-20-epoch-post_title

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: 2.5249
• Accuracy: 0.6975

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	0.7856	0.6475
No log	2.0	400	0.7447	0.71
0.7757	3.0	600	0.7684	0.6925
0.7757	4.0	800	0.9055	0.705
0.3775	5.0	1000	0.9552	0.715
0.3775	6.0	1200	1.3111	0.69
0.3775	7.0	1400	1.5227	0.715
0.1375	8.0	1600	1.8561	0.6975
0.1375	9.0	1800	1.9567	0.7
0.0543	10.0	2000	2.2423	0.6875
0.0543	11.0	2200	2.0781	0.695
0.0543	12.0	2400	2.3912	0.6725
0.0273	13.0	2600	2.3334	0.6975
0.0273	14.0	2800	2.3251	0.715
0.0144	15.0	3000	2.5461	0.6875
0.0144	16.0	3200	2.5073	0.71
0.0144	17.0	3400	2.4647	0.6925
0.0079	18.0	3600	2.5586	0.715
0.0079	19.0	3800	2.5277	0.705
0.005	20.0	4000	2.5249	0.6975

Framework versions

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