bertweet-fake-news-detector-TEST2

This model is a fine-tuned version of vinai/bertweet-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2043
• Accuracy: 0.9462
• Precision: 0.9504
• Recall: 0.9454
• F1: 0.9479
• Confusion Matrix: [[18414, 1028], [1136, 19682]]

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: 4e-05
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 128
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1	Confusion Matrix
0.1974	1.0	661	0.2095	0.9412	0.9361	0.9508	0.9434	[[4240, 314], [238, 4602]]
0.2155	2.0	1322	0.2063	0.9432	0.9378	0.9529	0.9453	[[4248, 306], [228, 4612]]
0.1414	3.0	1983	0.2040	0.9468	0.9496	0.9469	0.9483	[[4311, 243], [257, 4583]]
0.1708	3.9947	2640	0.2037	0.9460	0.9470	0.9483	0.9477	[[4297, 257], [250, 4590]]

Framework versions

• Transformers 4.51.3
• Pytorch 2.5.1
• Datasets 3.5.0
• Tokenizers 0.21.1