robeczech_lr1e-05_bs16_train150

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: 0.3536
• Precision: 0.8941
• Recall: 0.9034
• F1: 0.8987
• Accuracy: 0.9477

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: 1e-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	10	1.8823	0.0	0.0	0.0	0.5666
No log	2.0	20	1.6416	0.0	0.0	0.0	0.5666
No log	3.0	30	1.4591	0.0	0.0	0.0	0.5666
No log	4.0	40	1.2836	0.375	0.0217	0.0411	0.5760
No log	5.0	50	1.1520	0.4166	0.2665	0.3251	0.6794
No log	6.0	60	1.0409	0.4572	0.4070	0.4307	0.7363
No log	7.0	70	0.9407	0.6091	0.5794	0.5939	0.8093
No log	8.0	80	0.8449	0.6839	0.6779	0.6809	0.8485
No log	9.0	90	0.7620	0.7035	0.7079	0.7057	0.8602
No log	10.0	100	0.6857	0.7370	0.7349	0.7360	0.8738
No log	11.0	110	0.6316	0.7570	0.7610	0.7590	0.8845
No log	12.0	120	0.5793	0.7782	0.7793	0.7788	0.8937
No log	13.0	130	0.5451	0.7977	0.7996	0.7986	0.9016
No log	14.0	140	0.5131	0.8194	0.8238	0.8216	0.9127
No log	15.0	150	0.4870	0.8508	0.8479	0.8493	0.9253
No log	16.0	160	0.4638	0.8601	0.8580	0.8591	0.9297
No log	17.0	170	0.4476	0.8615	0.8682	0.8648	0.9316
No log	18.0	180	0.4290	0.8755	0.8730	0.8743	0.9364
No log	19.0	190	0.4160	0.8792	0.8817	0.8804	0.9391
No log	20.0	200	0.4037	0.8841	0.8841	0.8841	0.9410
No log	21.0	210	0.3962	0.8850	0.8885	0.8867	0.9418
No log	22.0	220	0.3911	0.8831	0.8865	0.8848	0.9404
No log	23.0	230	0.3850	0.8850	0.8880	0.8865	0.9416
No log	24.0	240	0.3817	0.8869	0.8827	0.8848	0.9408
No log	25.0	250	0.3764	0.8879	0.8875	0.8877	0.9424
No log	26.0	260	0.3730	0.8857	0.8870	0.8864	0.9416
No log	27.0	270	0.3691	0.8888	0.8880	0.8884	0.9427
No log	28.0	280	0.3663	0.8877	0.8889	0.8883	0.9427
No log	29.0	290	0.3657	0.8881	0.8894	0.8888	0.9429
No log	30.0	300	0.3658	0.8880	0.8885	0.8882	0.9427

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.1