czert_lr2e-05_bs4_train287_label_subtokens_False

This model is a fine-tuned version of UWB-AIR/Czert-B-base-cased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1569
• Precision: 0.9242
• Recall: 0.9210
• F1: 0.9226
• Accuracy: 0.9552

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: 4
• eval_batch_size: 4
• 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: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	72	0.2919	0.8337	0.8353	0.8345	0.9127
No log	2.0	144	0.1839	0.9057	0.8909	0.8982	0.9443
No log	3.0	216	0.1741	0.9205	0.8943	0.9072	0.9489
No log	4.0	288	0.1533	0.9293	0.9208	0.9251	0.9598
No log	5.0	360	0.1462	0.9328	0.9116	0.9221	0.9565
No log	6.0	432	0.1528	0.9304	0.9232	0.9268	0.9598
0.1893	7.0	504	0.1633	0.9303	0.9218	0.9260	0.9594
0.1893	8.0	576	0.1645	0.9290	0.9223	0.9256	0.9596

Framework versions

• Transformers 4.51.3
• Pytorch 2.7.0+cu126
• Datasets 3.6.0
• Tokenizers 0.21.1