deberta-v3-base-clickbait-task1-20-epoch-post_title

This model is a fine-tuned version of microsoft/deberta-v3-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.4262
• 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	0.7682	0.715
No log	2.0	400	0.7289	0.7125
0.7821	3.0	600	0.7968	0.705
0.7821	4.0	800	0.9608	0.73
0.3436	5.0	1000	1.0204	0.7275
0.3436	6.0	1200	1.3033	0.71
0.3436	7.0	1400	1.4924	0.7325
0.1188	8.0	1600	1.6538	0.7325
0.1188	9.0	1800	2.1675	0.7
0.0458	10.0	2000	2.0062	0.71
0.0458	11.0	2200	2.1166	0.7
0.0458	12.0	2400	2.3641	0.71
0.0147	13.0	2600	2.3338	0.7
0.0147	14.0	2800	2.2134	0.7125
0.0106	15.0	3000	2.3733	0.72
0.0106	16.0	3200	2.3268	0.7175
0.0106	17.0	3400	2.3951	0.7125
0.0041	18.0	3600	2.4211	0.7275
0.0041	19.0	3800	2.4168	0.725
0.0046	20.0	4000	2.4262	0.7275

Framework versions

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