Robeczech-2

This model is a fine-tuned version of ufal/robeczech-base on the Czech Relationship Extraction Dataset (http://hdl.handle.net/11234/1-3265). It achieves the following results on the evaluation set:

  • Loss: 0.8211
  • Accuracy: 0.8893
  • Micro Precision: 0.8893
  • Micro Recall: 0.8893
  • Micro F1: 0.8893
  • Macro Precision: 0.8687
  • Macro Recall: 0.8432
  • Macro F1: 0.8512

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: 5e-05
  • train_batch_size: 12
  • eval_batch_size: 12
  • seed: 42
  • 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: 8

Training results

Training Loss Epoch Step Validation Loss Accuracy Micro Precision Micro Recall Micro F1 Macro Precision Macro Recall Macro F1
0.5345 1.0 30148 0.5007 0.8546 0.8546 0.8546 0.8546 0.8145 0.7847 0.7893
0.4516 2.0 60296 0.4827 0.8759 0.8759 0.8759 0.8759 0.8478 0.8281 0.8293
0.3227 3.0 90444 0.4958 0.8839 0.8839 0.8839 0.8839 0.8564 0.8298 0.8357
0.2641 4.0 120592 0.5288 0.8850 0.8850 0.8850 0.8850 0.8506 0.8405 0.8411
0.2054 5.0 150740 0.6073 0.8862 0.8862 0.8862 0.8862 0.8487 0.8435 0.8419
0.1454 6.0 180888 0.6945 0.8862 0.8862 0.8862 0.8862 0.8545 0.8399 0.8415
0.0747 7.0 211036 0.7675 0.8926 0.8926 0.8926 0.8926 0.8584 0.8525 0.8516
0.0472 8.0 241184 0.8169 0.8917 0.8917 0.8917 0.8917 0.8561 0.8471 0.8477

Framework versions

  • Transformers 4.46.2
  • Pytorch 2.5.1+cu124
  • Datasets 3.1.0
  • Tokenizers 0.20.3
Downloads last month
51
Safetensors
Model size
126M params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for stulcrad/Robeczech-CERED2

Finetuned
(7)
this model

Dataset used to train stulcrad/Robeczech-CERED2