twitter-roberta-base-sentiment-latest-clickbait-task1-20-epoch-post

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

• Loss: 2.6694
• Accuracy: 0.7175

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	0.7795	0.655
No log	2.0	400	0.8334	0.6925
0.7177	3.0	600	0.8965	0.6925
0.7177	4.0	800	1.0902	0.7075
0.272	5.0	1000	1.3332	0.6875
0.272	6.0	1200	1.7338	0.6925
0.272	7.0	1400	2.0445	0.675
0.0924	8.0	1600	2.1525	0.7
0.0924	9.0	1800	2.2859	0.69
0.0515	10.0	2000	2.3061	0.7175
0.0515	11.0	2200	2.3235	0.71
0.0515	12.0	2400	2.5036	0.69
0.0213	13.0	2600	2.5110	0.7
0.0213	14.0	2800	2.5487	0.7
0.0162	15.0	3000	2.5632	0.7125
0.0162	16.0	3200	2.5690	0.7075
0.0162	17.0	3400	2.6913	0.685
0.0085	18.0	3600	2.7971	0.7
0.0085	19.0	3800	2.7057	0.715
0.0048	20.0	4000	2.6694	0.7175

Framework versions

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