This model is a fine-tuned version of microsoft/layoutlmv2-base-uncased on the funsd dataset. It achieves the following results on the evaluation set:

Loss: 1.6316
Answer: {'precision': 0.06269757639620653, 'recall': 0.14709517923362175, 'f1': 0.08792020687107498, 'number': 809}
Header: {'precision': 0.02142857142857143, 'recall': 0.025210084033613446, 'f1': 0.023166023166023165, 'number': 119}
Question: {'precision': 0.17976470588235294, 'recall': 0.3586854460093897, 'f1': 0.2394984326018809, 'number': 1065}
Overall Precision: 0.1211
Overall Recall: 0.2529
Overall F1: 0.1637
Overall Accuracy: 0.3969

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Answer	Header	Question	Overall Precision	Overall Recall	Overall F1	Overall Accuracy
1.9011	1.0	10	1.8281	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 809}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 119}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 1065}	0.0	0.0	0.0	0.2901
1.7212	2.0	20	1.6755	{'precision': 0.010714285714285714, 'recall': 0.011124845488257108, 'f1': 0.01091570648878108, 'number': 809}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 119}	{'precision': 0.8, 'recall': 0.003755868544600939, 'f1': 0.007476635514018691, 'number': 1065}	0.0154	0.0065	0.0092	0.3484
1.6644	3.0	30	1.7453	{'precision': 0.0030959752321981426, 'recall': 0.0012360939431396785, 'f1': 0.0017667844522968195, 'number': 809}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 119}	{'precision': 0.4895833333333333, 'recall': 0.044131455399061034, 'f1': 0.08096468561584841, 'number': 1065}	0.1143	0.0241	0.0398	0.3130
1.5949	4.0	40	1.7670	{'precision': 0.03335250143760782, 'recall': 0.07169344870210136, 'f1': 0.04552590266875981, 'number': 809}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 119}	{'precision': 0.13106796116504854, 'recall': 0.2535211267605634, 'f1': 0.1728, 'number': 1065}	0.0859	0.1646	0.1129	0.3285
1.4559	5.0	50	1.5921	{'precision': 0.05108940646130729, 'recall': 0.08405438813349815, 'f1': 0.06355140186915888, 'number': 809}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 119}	{'precision': 0.17631224764468373, 'recall': 0.2460093896713615, 'f1': 0.20540964327714623, 'number': 1065}	0.1171	0.1656	0.1372	0.3763
1.3707	6.0	60	1.6238	{'precision': 0.049044914816726896, 'recall': 0.11742892459826947, 'f1': 0.06919155134741442, 'number': 809}	{'precision': 0.006622516556291391, 'recall': 0.008403361344537815, 'f1': 0.007407407407407407, 'number': 119}	{'precision': 0.16497339138848574, 'recall': 0.32018779342723, 'f1': 0.21775223499361432, 'number': 1065}	0.1052	0.2193	0.1422	0.3786
1.2836	7.0	70	1.5846	{'precision': 0.058695652173913045, 'recall': 0.1334981458590853, 'f1': 0.08154020385050964, 'number': 809}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 119}	{'precision': 0.15556492411467115, 'recall': 0.3464788732394366, 'f1': 0.2147221414023858, 'number': 1065}	0.1120	0.2393	0.1526	0.3851
1.2161	8.0	80	1.6814	{'precision': 0.06025974025974026, 'recall': 0.1433868974042027, 'f1': 0.08485735186539868, 'number': 809}	{'precision': 0.009433962264150943, 'recall': 0.008403361344537815, 'f1': 0.008888888888888889, 'number': 119}	{'precision': 0.1631912964641886, 'recall': 0.3380281690140845, 'f1': 0.22011617242433507, 'number': 1065}	0.1126	0.2393	0.1531	0.3730
1.1499	9.0	90	1.6027	{'precision': 0.06253521126760564, 'recall': 0.13720642768850433, 'f1': 0.08591331269349846, 'number': 809}	{'precision': 0.011111111111111112, 'recall': 0.008403361344537815, 'f1': 0.009569377990430622, 'number': 119}	{'precision': 0.1547870097005483, 'recall': 0.34460093896713617, 'f1': 0.21362048894062866, 'number': 1065}	0.1131	0.2403	0.1538	0.3710
1.1199	10.0	100	1.6616	{'precision': 0.055350553505535055, 'recall': 0.12978986402966625, 'f1': 0.07760532150776053, 'number': 809}	{'precision': 0.019867549668874173, 'recall': 0.025210084033613446, 'f1': 0.022222222222222223, 'number': 119}	{'precision': 0.16207042851081885, 'recall': 0.3586854460093897, 'f1': 0.22326125073056693, 'number': 1065}	0.1112	0.2459	0.1532	0.3719
1.0651	11.0	110	1.6100	{'precision': 0.06031016657093624, 'recall': 0.12978986402966625, 'f1': 0.08235294117647059, 'number': 809}	{'precision': 0.015151515151515152, 'recall': 0.01680672268907563, 'f1': 0.01593625498007968, 'number': 119}	{'precision': 0.163854351687389, 'recall': 0.3464788732394366, 'f1': 0.22249020198974978, 'number': 1065}	0.1154	0.2388	0.1556	0.3805
1.0454	12.0	120	1.5988	{'precision': 0.0639269406392694, 'recall': 0.138442521631644, 'f1': 0.08746583365872705, 'number': 809}	{'precision': 0.022222222222222223, 'recall': 0.025210084033613446, 'f1': 0.02362204724409449, 'number': 119}	{'precision': 0.17867298578199053, 'recall': 0.3539906103286385, 'f1': 0.23748031496062993, 'number': 1065}	0.1231	0.2469	0.1643	0.3977
1.0279	13.0	130	1.6209	{'precision': 0.06463104325699745, 'recall': 0.15698393077873918, 'f1': 0.09156452775775054, 'number': 809}	{'precision': 0.022388059701492536, 'recall': 0.025210084033613446, 'f1': 0.02371541501976284, 'number': 119}	{'precision': 0.18118811881188118, 'recall': 0.3436619718309859, 'f1': 0.23727714748784443, 'number': 1065}	0.1204	0.2489	0.1623	0.3998
1.008	14.0	140	1.6538	{'precision': 0.0633116883116883, 'recall': 0.1446229913473424, 'f1': 0.08806925103500188, 'number': 809}	{'precision': 0.022556390977443608, 'recall': 0.025210084033613446, 'f1': 0.023809523809523808, 'number': 119}	{'precision': 0.18536350505536833, 'recall': 0.3615023474178404, 'f1': 0.24506683640992996, 'number': 1065}	0.1244	0.2534	0.1669	0.3965
0.9812	15.0	150	1.6316	{'precision': 0.06269757639620653, 'recall': 0.14709517923362175, 'f1': 0.08792020687107498, 'number': 809}	{'precision': 0.02142857142857143, 'recall': 0.025210084033613446, 'f1': 0.023166023166023165, 'number': 119}	{'precision': 0.17976470588235294, 'recall': 0.3586854460093897, 'f1': 0.2394984326018809, 'number': 1065}	0.1211	0.2529	0.1637	0.3969

Framework versions

Transformers 4.32.0
Pytorch 2.0.0+cu118
Datasets 2.17.1
Tokenizers 0.13.2

pedro1111/layoutkv

layoutkv

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

pedro1111/
layoutkv