BERTInvoiceCzech

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

• Loss: 0.0222
• Precision: 0.9591
• Recall: 0.9633
• F1: 0.9612
• Accuracy: 0.9929

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: 2
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	42	1.2228	0.0	0.0	0.0	0.7831
No log	2.0	84	0.5729	0.4183	0.4014	0.4097	0.8535
No log	3.0	126	0.3043	0.6654	0.6861	0.6756	0.9155
No log	4.0	168	0.2019	0.7618	0.7872	0.7743	0.9406
No log	5.0	210	0.1278	0.8161	0.8587	0.8369	0.9627
No log	6.0	252	0.0823	0.8727	0.9063	0.8892	0.9762
No log	7.0	294	0.0599	0.9086	0.9284	0.9184	0.9824
No log	8.0	336	0.0451	0.9335	0.9484	0.9409	0.9864
No log	9.0	378	0.0373	0.9388	0.9499	0.9443	0.9877
No log	10.0	420	0.0323	0.9458	0.9558	0.9508	0.9897
No log	11.0	462	0.0283	0.9506	0.9580	0.9543	0.9914
0.4073	12.0	504	0.0277	0.9567	0.9620	0.9594	0.9920
0.4073	13.0	546	0.0243	0.9517	0.9568	0.9542	0.9916
0.4073	14.0	588	0.0256	0.9610	0.9661	0.9635	0.9928
0.4073	15.0	630	0.0245	0.9588	0.9633	0.9610	0.9927
0.4073	16.0	672	0.0231	0.9606	0.9636	0.9621	0.9930
0.4073	17.0	714	0.0239	0.9582	0.9627	0.9604	0.9925
0.4073	18.0	756	0.0221	0.9606	0.9642	0.9624	0.9931
0.4073	19.0	798	0.0222	0.9594	0.9639	0.9617	0.9930
0.4073	20.0	840	0.0222	0.9591	0.9633	0.9612	0.9929

Framework versions

• Transformers 4.57.6
• Pytorch 2.9.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.2