Albanian-NER-mod

• Loss: 0.0811
• F1: 0.8808
• Report: precision recall f1-score support

    BDAY       0.98      1.00      0.99        52
     LOC       0.66      0.72      0.69        72
     ORG       0.78      0.73      0.75        85
     PER       0.93      0.94      0.94       355

  micro avg 0.88 0.88 0.88 564 macro avg 0.84 0.85 0.84 564 weighted avg 0.88 0.88 0.88 564

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

• learning_rate: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• 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: 2
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

    BDAY       0.98      0.98      0.98        52
     LOC       0.70      0.54      0.61        72
     ORG       0.59      0.59      0.59        85
     PER       0.75      0.90      0.82       355

    BDAY       0.98      0.98      0.98        52
     LOC       0.59      0.65      0.62        72
     ORG       0.67      0.64      0.65        85
     PER       0.94      0.90      0.92       355

    BDAY       0.96      1.00      0.98        52
     LOC       0.60      0.69      0.65        72
     ORG       0.70      0.73      0.71        85
     PER       0.92      0.94      0.93       355

    BDAY       0.98      1.00      0.99        52
     LOC       0.66      0.69      0.68        72
     ORG       0.72      0.71      0.71        85
     PER       0.92      0.95      0.93       355

    BDAY       0.98      1.00      0.99        52
     LOC       0.66      0.72      0.69        72
     ORG       0.78      0.73      0.75        85
     PER       0.93      0.94      0.94       355

Framework versions

• Transformers 5.0.0
• Pytorch 2.10.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.2