husst-hubert-hungarian

This model is a fine-tuned version of SZTAKI-HLT/hubert-base-cc on ariel-ml/HuSST-augmented dataset.

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• 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
• num_epochs: 4

Results

It achieves the following results on the evaluation set:

              precision    recall  f1-score   support

           0       0.77      0.90      0.83       697
           1       0.79      0.54      0.64       435
           2       0.45      0.67      0.54        33

    accuracy                           0.76      1165
   macro avg       0.67      0.70      0.67      1165
weighted avg       0.77      0.76      0.75      1165

Framework versions

• Transformers 4.48.0
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0