CNEC_2_0_ext_robeczech-base

This model is a fine-tuned version of ufal/robeczech-base on the cnec dataset. It achieves the following results on the evaluation set:

• Loss: 0.1663
• Precision: 0.8633
• Recall: 0.8933
• F1: 0.8780
• Accuracy: 0.9703

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: 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: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.2593	4.46	1000	0.1653	0.8195	0.8223	0.8209	0.9593
0.1209	8.93	2000	0.1355	0.8441	0.8789	0.8612	0.9679
0.0763	13.39	3000	0.1310	0.8591	0.8893	0.8739	0.9709
0.0539	17.86	4000	0.1383	0.8656	0.8953	0.8802	0.9719
0.0403	22.32	5000	0.1392	0.8626	0.8943	0.8782	0.9710
0.0316	26.79	6000	0.1539	0.8606	0.8948	0.8774	0.9712
0.0254	31.25	7000	0.1552	0.8660	0.8913	0.8785	0.9706
0.0211	35.71	8000	0.1621	0.8658	0.8968	0.8810	0.9701
0.0183	40.18	9000	0.1593	0.8688	0.8973	0.8828	0.9718
0.0161	44.64	10000	0.1638	0.8653	0.8993	0.8820	0.9714
0.015	49.11	11000	0.1663	0.8633	0.8933	0.8780	0.9703

Framework versions

• Transformers 4.36.2
• Pytorch 2.1.2+cu121
• Datasets 2.16.1
• Tokenizers 0.15.0