Layoutlmv3InvoiceCzechV0123Test

This model is a fine-tuned version of TomasFAV/Layoutlmv3InvoiceCzechV012 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0422
• Precision: 0.9203
• Recall: 0.9374
• F1: 0.9288
• Accuracy: 0.9928

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: 8
• eval_batch_size: 1
• 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_steps: 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	23	0.0638	0.8128	0.8376	0.825	0.9834
No log	2.0	46	0.0494	0.8666	0.8680	0.8673	0.9874
No log	3.0	69	0.0400	0.8876	0.8951	0.8913	0.9905
No log	4.0	92	0.0338	0.9048	0.9323	0.9183	0.9926
No log	5.0	115	0.0362	0.9021	0.9357	0.9186	0.9924
No log	6.0	138	0.0365	0.9150	0.9103	0.9126	0.9922
No log	7.0	161	0.0382	0.9026	0.9408	0.9213	0.9923
No log	8.0	184	0.0404	0.9003	0.9171	0.9086	0.9922
No log	9.0	207	0.0375	0.9121	0.9306	0.9213	0.9925
No log	10.0	230	0.0439	0.9115	0.9239	0.9176	0.9921
No log	11.0	253	0.0399	0.8961	0.9340	0.9147	0.9918
No log	12.0	276	0.0424	0.9153	0.9323	0.9237	0.9926
No log	13.0	299	0.0429	0.9091	0.9137	0.9114	0.9922
No log	14.0	322	0.0434	0.9274	0.9289	0.9281	0.9925
No log	15.0	345	0.0433	0.9193	0.9255	0.9224	0.9925
No log	16.0	368	0.0440	0.9174	0.9205	0.9189	0.9923
No log	17.0	391	0.0422	0.9203	0.9374	0.9288	0.9928
No log	18.0	414	0.0429	0.9138	0.9323	0.9229	0.9925
No log	19.0	437	0.0434	0.9164	0.9272	0.9218	0.9925
No log	20.0	460	0.0430	0.9133	0.9272	0.9202	0.9924

Framework versions

• Transformers 5.0.0
• Pytorch 2.10.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.2