Edit model card

lilt-en-funsd

This model is a fine-tuned version of SCUT-DLVCLab/lilt-roberta-en-base on the funsd-layoutlmv3 dataset. It achieves the following results on the evaluation set:

  • Loss: 1.7487
  • Answer: {'precision': 0.8851674641148325, 'recall': 0.9057527539779682, 'f1': 0.8953418027828192, 'number': 817}
  • Header: {'precision': 0.65625, 'recall': 0.5294117647058824, 'f1': 0.586046511627907, 'number': 119}
  • Question: {'precision': 0.8825088339222615, 'recall': 0.9275766016713092, 'f1': 0.9044816659121775, 'number': 1077}
  • Overall Precision: 0.8731
  • Overall Recall: 0.8952
  • Overall F1: 0.8840
  • Overall Accuracy: 0.7977

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: 8
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • training_steps: 2500

Training results

Training Loss Epoch Step Validation Loss Answer Header Question Overall Precision Overall Recall Overall F1 Overall Accuracy
0.4352 10.53 200 0.9574 {'precision': 0.8385167464114832, 'recall': 0.8580171358629131, 'f1': 0.8481548699334543, 'number': 817} {'precision': 0.5673076923076923, 'recall': 0.4957983193277311, 'f1': 0.5291479820627802, 'number': 119} {'precision': 0.8394534585824082, 'recall': 0.9127205199628597, 'f1': 0.8745551601423487, 'number': 1077} 0.8257 0.8659 0.8453 0.7896
0.0467 21.05 400 1.3446 {'precision': 0.8343057176196033, 'recall': 0.8751529987760098, 'f1': 0.8542413381123058, 'number': 817} {'precision': 0.5789473684210527, 'recall': 0.46218487394957986, 'f1': 0.514018691588785, 'number': 119} {'precision': 0.8543859649122807, 'recall': 0.904363974001857, 'f1': 0.8786648624267027, 'number': 1077} 0.8337 0.8664 0.8497 0.7969
0.0125 31.58 600 1.4274 {'precision': 0.8556461001164144, 'recall': 0.8996328029375765, 'f1': 0.8770883054892601, 'number': 817} {'precision': 0.568, 'recall': 0.5966386554621849, 'f1': 0.5819672131147541, 'number': 119} {'precision': 0.8916211293260473, 'recall': 0.9090064995357474, 'f1': 0.9002298850574713, 'number': 1077} 0.8573 0.8867 0.8718 0.8010
0.0071 42.11 800 1.4147 {'precision': 0.865265760197775, 'recall': 0.8567931456548348, 'f1': 0.8610086100861009, 'number': 817} {'precision': 0.6888888888888889, 'recall': 0.5210084033613446, 'f1': 0.5933014354066986, 'number': 119} {'precision': 0.8798206278026905, 'recall': 0.9108635097493036, 'f1': 0.8950729927007299, 'number': 1077} 0.8654 0.8659 0.8657 0.8055
0.0067 52.63 1000 1.5877 {'precision': 0.8747016706443914, 'recall': 0.8971848225214198, 'f1': 0.8858006042296073, 'number': 817} {'precision': 0.6074766355140186, 'recall': 0.5462184873949579, 'f1': 0.575221238938053, 'number': 119} {'precision': 0.8936363636363637, 'recall': 0.9127205199628597, 'f1': 0.9030776297657327, 'number': 1077} 0.8709 0.8847 0.8778 0.8080
0.003 63.16 1200 1.5406 {'precision': 0.875, 'recall': 0.8996328029375765, 'f1': 0.8871454435727218, 'number': 817} {'precision': 0.584070796460177, 'recall': 0.5546218487394958, 'f1': 0.5689655172413793, 'number': 119} {'precision': 0.8858695652173914, 'recall': 0.9080779944289693, 'f1': 0.8968363136176066, 'number': 1077} 0.8649 0.8838 0.8742 0.8183
0.0011 73.68 1400 1.6131 {'precision': 0.8686987104337632, 'recall': 0.9069767441860465, 'f1': 0.8874251497005988, 'number': 817} {'precision': 0.7011494252873564, 'recall': 0.5126050420168067, 'f1': 0.5922330097087377, 'number': 119} {'precision': 0.8821966341895483, 'recall': 0.924791086350975, 'f1': 0.9029918404351769, 'number': 1077} 0.8690 0.8932 0.8809 0.8111
0.0008 84.21 1600 1.7487 {'precision': 0.8851674641148325, 'recall': 0.9057527539779682, 'f1': 0.8953418027828192, 'number': 817} {'precision': 0.65625, 'recall': 0.5294117647058824, 'f1': 0.586046511627907, 'number': 119} {'precision': 0.8825088339222615, 'recall': 0.9275766016713092, 'f1': 0.9044816659121775, 'number': 1077} 0.8731 0.8952 0.8840 0.7977
0.0007 94.74 1800 1.8317 {'precision': 0.8605990783410138, 'recall': 0.9143206854345165, 'f1': 0.886646884272997, 'number': 817} {'precision': 0.6095238095238096, 'recall': 0.5378151260504201, 'f1': 0.5714285714285715, 'number': 119} {'precision': 0.9026876737720111, 'recall': 0.904363974001857, 'f1': 0.9035250463821892, 'number': 1077} 0.8699 0.8867 0.8782 0.7934
0.0005 105.26 2000 1.8600 {'precision': 0.8669778296382731, 'recall': 0.9094247246022031, 'f1': 0.8876941457586618, 'number': 817} {'precision': 0.6428571428571429, 'recall': 0.5294117647058824, 'f1': 0.5806451612903226, 'number': 119} {'precision': 0.8928247048138056, 'recall': 0.9127205199628597, 'f1': 0.9026629935720845, 'number': 1077} 0.8701 0.8887 0.8793 0.7897
0.0005 115.79 2200 1.7672 {'precision': 0.8781362007168458, 'recall': 0.8996328029375765, 'f1': 0.8887545344619106, 'number': 817} {'precision': 0.6568627450980392, 'recall': 0.5630252100840336, 'f1': 0.6063348416289592, 'number': 119} {'precision': 0.8939256572982774, 'recall': 0.9155060352831941, 'f1': 0.9045871559633027, 'number': 1077} 0.8756 0.8882 0.8819 0.8074
0.0002 126.32 2400 1.8044 {'precision': 0.8604382929642446, 'recall': 0.9130966952264382, 'f1': 0.8859857482185273, 'number': 817} {'precision': 0.6391752577319587, 'recall': 0.5210084033613446, 'f1': 0.5740740740740741, 'number': 119} {'precision': 0.9022140221402214, 'recall': 0.9080779944289693, 'f1': 0.9051365108745951, 'number': 1077} 0.8721 0.8872 0.8796 0.8000

Framework versions

  • Transformers 4.32.1
  • Pytorch 2.0.1+cu118
  • Datasets 2.14.4
  • Tokenizers 0.13.3
Downloads last month
8

Finetuned from