twitter-stance-detection-foldfold0

This model is a fine-tuned version of aubmindlab/bert-base-arabertv02-twitter on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2042
• Accuracy: 0.7956
• Macro F1: 0.7952
• Weighted F1: 0.7956
• F1 Pro: 0.8387
• F1 Against: 0.768
• F1 Neutral: 0.7788

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
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 0.1
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Macro F1	Weighted F1	F1 Pro	F1 Against	F1 Neutral
0.7083	1.8679	50	0.2878	0.7238	0.7137	0.7114	0.7910	0.5882	0.7619
0.3886	3.7170	100	0.2281	0.7459	0.7477	0.7485	0.8036	0.7211	0.7184
0.2775	5.5660	150	0.1970	0.7790	0.7814	0.7807	0.8070	0.7407	0.7965
0.1809	7.4151	200	0.2063	0.7845	0.7838	0.7846	0.8387	0.7556	0.7573
0.1379	9.2642	250	0.2042	0.7956	0.7952	0.7956	0.8387	0.768	0.7788

Framework versions

• Transformers 5.0.0
• Pytorch 2.9.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.2