robeczech_lr5e-05_bs16_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: 0.1534
• Precision: 0.9532
• Recall: 0.9571
• F1: 0.9552
• Accuracy: 0.9737

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: 16
• eval_batch_size: 16
• 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	18	1.0604	0.5553	0.4655	0.5064	0.7578
No log	2.0	36	0.5178	0.8104	0.8049	0.8077	0.9048
No log	3.0	54	0.3230	0.9009	0.9000	0.9005	0.9489
No log	4.0	72	0.2539	0.9277	0.9174	0.9226	0.9594
No log	5.0	90	0.2271	0.9369	0.9324	0.9347	0.9640
No log	6.0	108	0.2010	0.9332	0.9440	0.9386	0.9663
No log	7.0	126	0.1867	0.9425	0.9411	0.9418	0.9690
No log	8.0	144	0.1798	0.9402	0.9411	0.9406	0.9684
No log	9.0	162	0.1824	0.9411	0.9406	0.9408	0.9686
No log	10.0	180	0.1689	0.9500	0.9440	0.9470	0.9711
No log	11.0	198	0.1609	0.9547	0.9474	0.9510	0.9726
No log	12.0	216	0.1543	0.9542	0.9459	0.9500	0.9726
No log	13.0	234	0.1668	0.9503	0.9334	0.9418	0.9678

Framework versions

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