olid_training_better_bert_1

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

• Loss: 0.5048
• Accuracy Offensive: 0.8066
• F1 Offensive: 0.8090
• Accuracy Targeted: 0.7931
• F1 Targeted: 0.7831
• Accuracy Stance: 0.7591
• F1 Stance: 0.7455

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: 1e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy Offensive	F1 Offensive	Accuracy Targeted	F1 Targeted	Accuracy Stance	F1 Stance
No log	1.0	373	0.5088	0.8127	0.8061	0.7900	0.7704	0.7515	0.6998
0.5861	2.0	746	0.5051	0.8036	0.8070	0.7855	0.7777	0.7485	0.7190
0.4732	3.0	1119	0.5048	0.8066	0.8090	0.7931	0.7831	0.7591	0.7455
0.4732	4.0	1492	0.5172	0.8074	0.8083	0.7885	0.7784	0.7666	0.7534
0.4028	5.0	1865	0.6278	0.7628	0.7703	0.7424	0.7424	0.7054	0.7154

Framework versions

• Transformers 4.45.2
• Pytorch 2.5.0+cu124
• Datasets 3.0.1
• Tokenizers 0.20.1