metadata

license: mit
base_model: SCUT-DLVCLab/lilt-roberta-en-base
tags:
  - generated_from_trainer
datasets:
  - funsd-layoutlmv3
model-index:
  - name: lilt-en-funsd
    results: []

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: 0.5526
Answer: {'precision': 0.8434684684684685, 'recall': 0.9167686658506732, 'f1': 0.8785923753665689, 'number': 817}
Header: {'precision': 0.6494845360824743, 'recall': 0.5294117647058824, 'f1': 0.5833333333333334, 'number': 119}
Question: {'precision': 0.8929219600725953, 'recall': 0.9136490250696379, 'f1': 0.9031665901789812, 'number': 1077}
Overall Precision: 0.8606
Overall Recall: 0.8922
Overall F1: 0.8761
Overall Accuracy: 0.7988

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
0.1122	10.53	200	0.3663	{'precision': 0.7907488986784141, 'recall': 0.8788249694002448, 'f1': 0.832463768115942, 'number': 817}	{'precision': 0.4888888888888889, 'recall': 0.5546218487394958, 'f1': 0.5196850393700787, 'number': 119}	{'precision': 0.8885630498533724, 'recall': 0.8440111420612814, 'f1': 0.8657142857142858, 'number': 1077}	0.8195	0.8410	0.8301	0.7813
0.0121	21.05	400	0.3856	{'precision': 0.8256880733944955, 'recall': 0.8812729498164015, 'f1': 0.8525754884547069, 'number': 817}	{'precision': 0.6021505376344086, 'recall': 0.47058823529411764, 'f1': 0.5283018867924528, 'number': 119}	{'precision': 0.8739946380697051, 'recall': 0.9080779944289693, 'f1': 0.8907103825136613, 'number': 1077}	0.8417	0.8713	0.8562	0.8060
0.0038	31.58	600	0.4187	{'precision': 0.8352534562211982, 'recall': 0.8873929008567931, 'f1': 0.8605341246290802, 'number': 817}	{'precision': 0.59, 'recall': 0.4957983193277311, 'f1': 0.5388127853881278, 'number': 119}	{'precision': 0.8956197576887233, 'recall': 0.8922934076137419, 'f1': 0.8939534883720931, 'number': 1077}	0.8550	0.8669	0.8609	0.8082
0.0018	42.11	800	0.4984	{'precision': 0.8299319727891157, 'recall': 0.8959608323133414, 'f1': 0.8616833431430253, 'number': 817}	{'precision': 0.6111111111111112, 'recall': 0.46218487394957986, 'f1': 0.5263157894736842, 'number': 119}	{'precision': 0.8860055607043559, 'recall': 0.8876508820798514, 'f1': 0.8868274582560296, 'number': 1077}	0.8498	0.8659	0.8578	0.7950
0.0015	52.63	1000	0.4974	{'precision': 0.830316742081448, 'recall': 0.8984088127294981, 'f1': 0.8630217519106408, 'number': 817}	{'precision': 0.6161616161616161, 'recall': 0.5126050420168067, 'f1': 0.5596330275229358, 'number': 119}	{'precision': 0.9015151515151515, 'recall': 0.8839368616527391, 'f1': 0.8926394749179559, 'number': 1077}	0.8568	0.8679	0.8623	0.7888
0.0012	63.16	1200	0.5089	{'precision': 0.8566473988439306, 'recall': 0.9069767441860465, 'f1': 0.8810939357907253, 'number': 817}	{'precision': 0.5981308411214953, 'recall': 0.5378151260504201, 'f1': 0.5663716814159291, 'number': 119}	{'precision': 0.9059590316573557, 'recall': 0.903435468895079, 'f1': 0.904695490469549, 'number': 1077}	0.8690	0.8833	0.8761	0.8067
0.0006	73.68	1400	0.5012	{'precision': 0.8409090909090909, 'recall': 0.9057527539779682, 'f1': 0.8721272834413673, 'number': 817}	{'precision': 0.6043956043956044, 'recall': 0.46218487394957986, 'f1': 0.5238095238095237, 'number': 119}	{'precision': 0.8840182648401826, 'recall': 0.8987929433611885, 'f1': 0.8913443830570903, 'number': 1077}	0.8533	0.8758	0.8644	0.7997
0.0003	84.21	1600	0.5316	{'precision': 0.8506944444444444, 'recall': 0.8996328029375765, 'f1': 0.874479476502082, 'number': 817}	{'precision': 0.5775862068965517, 'recall': 0.5630252100840336, 'f1': 0.5702127659574467, 'number': 119}	{'precision': 0.8932481751824818, 'recall': 0.9090064995357474, 'f1': 0.9010584445467096, 'number': 1077}	0.8579	0.8847	0.8711	0.8003
0.0002	94.74	1800	0.5347	{'precision': 0.8617401668653158, 'recall': 0.8849449204406364, 'f1': 0.8731884057971013, 'number': 817}	{'precision': 0.5317460317460317, 'recall': 0.5630252100840336, 'f1': 0.5469387755102041, 'number': 119}	{'precision': 0.8999081726354453, 'recall': 0.9099350046425255, 'f1': 0.9048938134810711, 'number': 1077}	0.8617	0.8793	0.8704	0.7979
0.0002	105.26	2000	0.5526	{'precision': 0.8434684684684685, 'recall': 0.9167686658506732, 'f1': 0.8785923753665689, 'number': 817}	{'precision': 0.6494845360824743, 'recall': 0.5294117647058824, 'f1': 0.5833333333333334, 'number': 119}	{'precision': 0.8929219600725953, 'recall': 0.9136490250696379, 'f1': 0.9031665901789812, 'number': 1077}	0.8606	0.8922	0.8761	0.7988
0.0001	115.79	2200	0.5488	{'precision': 0.8368953880764904, 'recall': 0.9106487148102815, 'f1': 0.8722157092614302, 'number': 817}	{'precision': 0.5833333333333334, 'recall': 0.5294117647058824, 'f1': 0.5550660792951542, 'number': 119}	{'precision': 0.8835740072202166, 'recall': 0.9090064995357474, 'f1': 0.8961098398169337, 'number': 1077}	0.8485	0.8872	0.8674	0.8017
0.0001	126.32	2400	0.5510	{'precision': 0.8372615039281706, 'recall': 0.9130966952264382, 'f1': 0.8735362997658079, 'number': 817}	{'precision': 0.6074766355140186, 'recall': 0.5462184873949579, 'f1': 0.575221238938053, 'number': 119}	{'precision': 0.895264116575592, 'recall': 0.9127205199628597, 'f1': 0.9039080459770115, 'number': 1077}	0.8559	0.8912	0.8732	0.8024

Framework versions

Transformers 4.35.0
Pytorch 2.1.0
Datasets 2.9.0
Tokenizers 0.14.1