robeczech_lr3e-05_bs4_train287

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

• Loss: 0.1341
• Precision: 0.9463
• Recall: 0.9666
• F1: 0.9563
• Accuracy: 0.9734

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: 3e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Use OptimizerNames.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	0.4864	0.7926	0.8194	0.8058	0.9027
No log	2.0	144	0.2842	0.8967	0.8967	0.8967	0.9479
No log	3.0	216	0.2135	0.9343	0.9334	0.9338	0.9636
No log	4.0	288	0.1866	0.9367	0.9363	0.9365	0.9648
No log	5.0	360	0.1668	0.9434	0.9416	0.9425	0.9678
No log	6.0	432	0.1697	0.9454	0.9440	0.9447	0.9694
0.3144	7.0	504	0.1533	0.9450	0.9450	0.9450	0.9699
0.3144	8.0	576	0.1412	0.9498	0.9507	0.9503	0.9732
0.3144	9.0	648	0.1400	0.9513	0.9517	0.9515	0.9734
0.3144	10.0	720	0.1243	0.9500	0.9536	0.9518	0.9747
0.3144	11.0	792	0.1395	0.9526	0.9416	0.9471	0.9711
0.3144	12.0	864	0.1492	0.9512	0.9416	0.9464	0.9709
0.3144	13.0	936	0.1378	0.9557	0.9488	0.9523	0.9734

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1