layoutlm-FUNSDxSynthetic-5fold

This model is a fine-tuned version of microsoft/layoutlm-base-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0137
• Eader: {'precision': 0.971830985915493, 'recall': 0.971830985915493, 'f1': 0.971830985915493, 'number': 71}
• Nswer: {'precision': 0.9921568627450981, 'recall': 0.98828125, 'f1': 0.990215264187867, 'number': 256}
• Uestion: {'precision': 0.9818181818181818, 'recall': 0.989010989010989, 'f1': 0.9854014598540145, 'number': 273}
• Overall Precision: 0.9850
• Overall Recall: 0.9867
• Overall F1: 0.9858
• Overall Accuracy: 0.9965

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: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Use 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: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Eader	Nswer	Uestion	Overall Precision	Overall Recall	Overall F1	Overall Accuracy
0.0504	1.0	11	0.0183	{'precision': 0.9722222222222222, 'recall': 0.9859154929577465, 'f1': 0.979020979020979, 'number': 71}	{'precision': 0.984375, 'recall': 0.984375, 'f1': 0.984375, 'number': 256}	{'precision': 0.9782608695652174, 'recall': 0.989010989010989, 'f1': 0.9836065573770493, 'number': 273}	0.9801	0.9867	0.9834	0.9956
0.0459	2.0	22	0.0175	{'precision': 0.971830985915493, 'recall': 0.971830985915493, 'f1': 0.971830985915493, 'number': 71}	{'precision': 0.98828125, 'recall': 0.98828125, 'f1': 0.98828125, 'number': 256}	{'precision': 0.9747292418772563, 'recall': 0.989010989010989, 'f1': 0.9818181818181818, 'number': 273}	0.9801	0.9867	0.9834	0.9953
0.0289	3.0	33	0.0146	{'precision': 0.9859154929577465, 'recall': 0.9859154929577465, 'f1': 0.9859154929577465, 'number': 71}	{'precision': 0.9921568627450981, 'recall': 0.98828125, 'f1': 0.990215264187867, 'number': 256}	{'precision': 0.9890909090909091, 'recall': 0.9963369963369964, 'f1': 0.9927007299270072, 'number': 273}	0.9900	0.9917	0.9908	0.9965
0.0243	4.0	44	0.0157	{'precision': 0.971830985915493, 'recall': 0.971830985915493, 'f1': 0.971830985915493, 'number': 71}	{'precision': 0.9921568627450981, 'recall': 0.98828125, 'f1': 0.990215264187867, 'number': 256}	{'precision': 0.9782608695652174, 'recall': 0.989010989010989, 'f1': 0.9836065573770493, 'number': 273}	0.9834	0.9867	0.9850	0.9958
0.0215	5.0	55	0.0129	{'precision': 0.971830985915493, 'recall': 0.971830985915493, 'f1': 0.971830985915493, 'number': 71}	{'precision': 0.9921875, 'recall': 0.9921875, 'f1': 0.9921875, 'number': 256}	{'precision': 0.9818181818181818, 'recall': 0.989010989010989, 'f1': 0.9854014598540145, 'number': 273}	0.9850	0.9883	0.9867	0.9975
0.0176	6.0	66	0.0151	{'precision': 0.971830985915493, 'recall': 0.971830985915493, 'f1': 0.971830985915493, 'number': 71}	{'precision': 0.9921875, 'recall': 0.9921875, 'f1': 0.9921875, 'number': 256}	{'precision': 0.9782608695652174, 'recall': 0.989010989010989, 'f1': 0.9836065573770493, 'number': 273}	0.9834	0.9883	0.9859	0.9963
0.0151	7.0	77	0.0149	{'precision': 0.9583333333333334, 'recall': 0.971830985915493, 'f1': 0.965034965034965, 'number': 71}	{'precision': 0.9921568627450981, 'recall': 0.98828125, 'f1': 0.990215264187867, 'number': 256}	{'precision': 0.9817518248175182, 'recall': 0.9853479853479854, 'f1': 0.9835466179159049, 'number': 273}	0.9834	0.985	0.9842	0.9968
0.0136	8.0	88	0.0142	{'precision': 0.9583333333333334, 'recall': 0.971830985915493, 'f1': 0.965034965034965, 'number': 71}	{'precision': 0.98828125, 'recall': 0.98828125, 'f1': 0.98828125, 'number': 256}	{'precision': 0.9781818181818182, 'recall': 0.9853479853479854, 'f1': 0.9817518248175183, 'number': 273}	0.9801	0.985	0.9825	0.9965
0.0136	9.0	99	0.0148	{'precision': 0.971830985915493, 'recall': 0.971830985915493, 'f1': 0.971830985915493, 'number': 71}	{'precision': 0.9921568627450981, 'recall': 0.98828125, 'f1': 0.990215264187867, 'number': 256}	{'precision': 0.9818181818181818, 'recall': 0.989010989010989, 'f1': 0.9854014598540145, 'number': 273}	0.9850	0.9867	0.9858	0.9963
0.0103	10.0	110	0.0138	{'precision': 0.9859154929577465, 'recall': 0.9859154929577465, 'f1': 0.9859154929577465, 'number': 71}	{'precision': 0.9921568627450981, 'recall': 0.98828125, 'f1': 0.990215264187867, 'number': 256}	{'precision': 0.9854545454545455, 'recall': 0.9926739926739927, 'f1': 0.989051094890511, 'number': 273}	0.9884	0.99	0.9892	0.9965
0.0091	11.0	121	0.0136	{'precision': 0.971830985915493, 'recall': 0.971830985915493, 'f1': 0.971830985915493, 'number': 71}	{'precision': 0.9921568627450981, 'recall': 0.98828125, 'f1': 0.990215264187867, 'number': 256}	{'precision': 0.9854014598540146, 'recall': 0.989010989010989, 'f1': 0.9872029250457038, 'number': 273}	0.9867	0.9867	0.9867	0.9968
0.0081	12.0	132	0.0131	{'precision': 0.9859154929577465, 'recall': 0.9859154929577465, 'f1': 0.9859154929577465, 'number': 71}	{'precision': 0.98828125, 'recall': 0.98828125, 'f1': 0.98828125, 'number': 256}	{'precision': 0.9890510948905109, 'recall': 0.9926739926739927, 'f1': 0.9908592321755026, 'number': 273}	0.9884	0.99	0.9892	0.9973
0.0098	13.0	143	0.0136	{'precision': 0.971830985915493, 'recall': 0.971830985915493, 'f1': 0.971830985915493, 'number': 71}	{'precision': 0.9921568627450981, 'recall': 0.98828125, 'f1': 0.990215264187867, 'number': 256}	{'precision': 0.9854014598540146, 'recall': 0.989010989010989, 'f1': 0.9872029250457038, 'number': 273}	0.9867	0.9867	0.9867	0.9968
0.0066	14.0	154	0.0139	{'precision': 0.971830985915493, 'recall': 0.971830985915493, 'f1': 0.971830985915493, 'number': 71}	{'precision': 0.9921568627450981, 'recall': 0.98828125, 'f1': 0.990215264187867, 'number': 256}	{'precision': 0.9818181818181818, 'recall': 0.989010989010989, 'f1': 0.9854014598540145, 'number': 273}	0.9850	0.9867	0.9858	0.9965
0.007	15.0	165	0.0137	{'precision': 0.971830985915493, 'recall': 0.971830985915493, 'f1': 0.971830985915493, 'number': 71}	{'precision': 0.9921568627450981, 'recall': 0.98828125, 'f1': 0.990215264187867, 'number': 256}	{'precision': 0.9818181818181818, 'recall': 0.989010989010989, 'f1': 0.9854014598540145, 'number': 273}	0.9850	0.9867	0.9858	0.9965

Framework versions

• Transformers 4.49.0
• Pytorch 2.6.0+cu124
• Datasets 3.3.2
• Tokenizers 0.21.0