layoutlm-base-uncased-finetuned-invoices-0

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: 0.0543
• B-adress: {'precision': 0.9086576648133439, 'recall': 0.967032967032967, 'f1': 0.9369369369369369, 'number': 1183}
• B-name: {'precision': 0.96045197740113, 'recall': 0.9941520467836257, 'f1': 0.9770114942528735, 'number': 342}
• Gst no: {'precision': 0.952755905511811, 'recall': 0.983739837398374, 'f1': 0.968, 'number': 123}
• Invoice no: {'precision': 0.94, 'recall': 0.94, 'f1': 0.94, 'number': 100}
• Order date: {'precision': 1.0, 'recall': 0.983739837398374, 'f1': 0.9918032786885246, 'number': 123}
• Order id: {'precision': 0.9922480620155039, 'recall': 0.9846153846153847, 'f1': 0.9884169884169884, 'number': 130}
• S-adress: {'precision': 0.9866666666666667, 'recall': 0.9458077709611452, 'f1': 0.9658052727747326, 'number': 1956}
• S-name: {'precision': 0.85431654676259, 'recall': 0.9875259875259875, 'f1': 0.9161041465766634, 'number': 481}
• Total gross: {'precision': 0.8545454545454545, 'recall': 0.8392857142857143, 'f1': 0.8468468468468467, 'number': 56}
• Total net: {'precision': 0.8768115942028986, 'recall': 0.952755905511811, 'f1': 0.9132075471698112, 'number': 127}
• Overall Precision: 0.9421
• Overall Recall: 0.9610
• Overall F1: 0.9515
• Overall Accuracy: 0.9872

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: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 10
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	B-adress	B-name	Gst no	Invoice no	Order date	Order id	S-adress	S-name	Total gross	Total net	Overall Precision	Overall Recall	Overall F1	Overall Accuracy
1.3404	1.0	19	0.5934	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 1183}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 342}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 123}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 100}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 123}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 130}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 1956}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 481}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 56}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 127}	0.0	0.0	0.0	0.8274
0.4824	2.0	38	0.3290	{'precision': 0.5835654596100278, 'recall': 0.35418427726120033, 'f1': 0.4408206207259337, 'number': 1183}	{'precision': 0.5, 'recall': 0.0029239766081871343, 'f1': 0.005813953488372093, 'number': 342}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 123}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 100}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 123}	{'precision': 1.0, 'recall': 0.4076923076923077, 'f1': 0.5792349726775957, 'number': 130}	{'precision': 0.8446411012782694, 'recall': 0.878323108384458, 'f1': 0.8611528822055139, 'number': 1956}	{'precision': 0.825, 'recall': 0.6860706860706861, 'f1': 0.7491486946651532, 'number': 481}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 56}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 127}	0.7861	0.5456	0.6441	0.9083
0.2716	3.0	57	0.2123	{'precision': 0.6143583227445998, 'recall': 0.8174133558748944, 'f1': 0.701487123685165, 'number': 1183}	{'precision': 0.856140350877193, 'recall': 0.7134502923976608, 'f1': 0.7783094098883573, 'number': 342}	{'precision': 0.8910891089108911, 'recall': 0.7317073170731707, 'f1': 0.8035714285714285, 'number': 123}	{'precision': 0.98, 'recall': 0.49, 'f1': 0.6533333333333333, 'number': 100}	{'precision': 0.9375, 'recall': 0.6097560975609756, 'f1': 0.7389162561576355, 'number': 123}	{'precision': 0.9223300970873787, 'recall': 0.7307692307692307, 'f1': 0.815450643776824, 'number': 130}	{'precision': 0.9909255898366606, 'recall': 0.8374233128834356, 'f1': 0.9077306733167083, 'number': 1956}	{'precision': 0.9159836065573771, 'recall': 0.9293139293139293, 'f1': 0.9226006191950464, 'number': 481}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 56}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 127}	0.8318	0.7801	0.8051	0.9439
0.1817	4.0	76	0.1527	{'precision': 0.7367706919945726, 'recall': 0.9180050718512257, 'f1': 0.8174633044787355, 'number': 1183}	{'precision': 0.8218085106382979, 'recall': 0.9035087719298246, 'f1': 0.8607242339832869, 'number': 342}	{'precision': 0.8928571428571429, 'recall': 0.8130081300813008, 'f1': 0.8510638297872342, 'number': 123}	{'precision': 0.8651685393258427, 'recall': 0.77, 'f1': 0.8148148148148148, 'number': 100}	{'precision': 0.9484536082474226, 'recall': 0.7479674796747967, 'f1': 0.8363636363636364, 'number': 123}	{'precision': 0.96, 'recall': 0.7384615384615385, 'f1': 0.8347826086956522, 'number': 130}	{'precision': 0.9706510138740662, 'recall': 0.929959100204499, 'f1': 0.9498694516971279, 'number': 1956}	{'precision': 0.9447852760736196, 'recall': 0.9604989604989606, 'f1': 0.9525773195876289, 'number': 481}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 56}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 127}	0.8764	0.8745	0.8754	0.9615
0.1376	5.0	95	0.1115	{'precision': 0.8572664359861591, 'recall': 0.8377007607776839, 'f1': 0.84737067122702, 'number': 1183}	{'precision': 0.8636363636363636, 'recall': 0.9444444444444444, 'f1': 0.9022346368715084, 'number': 342}	{'precision': 0.9363636363636364, 'recall': 0.8373983739837398, 'f1': 0.8841201716738197, 'number': 123}	{'precision': 0.875, 'recall': 0.84, 'f1': 0.8571428571428572, 'number': 100}	{'precision': 0.8899082568807339, 'recall': 0.7886178861788617, 'f1': 0.8362068965517241, 'number': 123}	{'precision': 0.9702970297029703, 'recall': 0.7538461538461538, 'f1': 0.8484848484848484, 'number': 130}	{'precision': 0.9693665628245067, 'recall': 0.9544989775051125, 'f1': 0.9618753219989695, 'number': 1956}	{'precision': 0.9665970772442589, 'recall': 0.9625779625779626, 'f1': 0.9645833333333335, 'number': 481}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 56}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 127}	0.9253	0.8712	0.8975	0.9683
0.1027	6.0	114	0.0891	{'precision': 0.8551068883610451, 'recall': 0.9129332206255283, 'f1': 0.8830744071954211, 'number': 1183}	{'precision': 0.9029649595687331, 'recall': 0.97953216374269, 'f1': 0.9396914446002805, 'number': 342}	{'precision': 0.9568965517241379, 'recall': 0.9024390243902439, 'f1': 0.9288702928870294, 'number': 123}	{'precision': 0.9052631578947369, 'recall': 0.86, 'f1': 0.8820512820512821, 'number': 100}	{'precision': 0.9067796610169492, 'recall': 0.8699186991869918, 'f1': 0.8879668049792531, 'number': 123}	{'precision': 0.9902912621359223, 'recall': 0.7846153846153846, 'f1': 0.8755364806866953, 'number': 130}	{'precision': 0.9579789894947474, 'recall': 0.9790388548057259, 'f1': 0.9683944374209861, 'number': 1956}	{'precision': 0.9747368421052631, 'recall': 0.9625779625779626, 'f1': 0.9686192468619247, 'number': 481}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 56}	{'precision': 0.6666666666666666, 'recall': 0.015748031496062992, 'f1': 0.03076923076923077, 'number': 127}	0.9247	0.9091	0.9168	0.9740
0.0774	7.0	133	0.0761	{'precision': 0.883495145631068, 'recall': 0.9230769230769231, 'f1': 0.9028524183546921, 'number': 1183}	{'precision': 0.9786585365853658, 'recall': 0.9385964912280702, 'f1': 0.9582089552238806, 'number': 342}	{'precision': 0.9576271186440678, 'recall': 0.9186991869918699, 'f1': 0.9377593360995852, 'number': 123}	{'precision': 0.9368421052631579, 'recall': 0.89, 'f1': 0.9128205128205129, 'number': 100}	{'precision': 0.9652173913043478, 'recall': 0.9024390243902439, 'f1': 0.9327731092436974, 'number': 123}	{'precision': 0.9908256880733946, 'recall': 0.8307692307692308, 'f1': 0.9037656903765691, 'number': 130}	{'precision': 0.9897739504843919, 'recall': 0.9401840490797546, 'f1': 0.9643418982695334, 'number': 1956}	{'precision': 0.9019230769230769, 'recall': 0.975051975051975, 'f1': 0.9370629370629371, 'number': 481}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 56}	{'precision': 0.5984251968503937, 'recall': 0.5984251968503937, 'f1': 0.5984251968503937, 'number': 127}	0.9361	0.9128	0.9243	0.9783
0.0604	8.0	152	0.0642	{'precision': 0.8949511400651465, 'recall': 0.9289940828402367, 'f1': 0.9116549149730402, 'number': 1183}	{'precision': 0.9598853868194842, 'recall': 0.97953216374269, 'f1': 0.9696092619392185, 'number': 342}	{'precision': 0.9603174603174603, 'recall': 0.983739837398374, 'f1': 0.9718875502008032, 'number': 123}	{'precision': 0.9468085106382979, 'recall': 0.89, 'f1': 0.9175257731958764, 'number': 100}	{'precision': 0.9910714285714286, 'recall': 0.9024390243902439, 'f1': 0.9446808510638298, 'number': 123}	{'precision': 1.0, 'recall': 0.8615384615384616, 'f1': 0.9256198347107438, 'number': 130}	{'precision': 0.9896061269146609, 'recall': 0.9248466257668712, 'f1': 0.9561310782241015, 'number': 1956}	{'precision': 0.8571428571428571, 'recall': 0.9854469854469855, 'f1': 0.9168278529980657, 'number': 481}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 56}	{'precision': 0.5363128491620112, 'recall': 0.7559055118110236, 'f1': 0.627450980392157, 'number': 127}	0.9269	0.9188	0.9228	0.9798
0.0472	9.0	171	0.0684	{'precision': 0.8824439288476411, 'recall': 0.9644970414201184, 'f1': 0.9216478190630049, 'number': 1183}	{'precision': 0.9629629629629629, 'recall': 0.9883040935672515, 'f1': 0.9754689754689754, 'number': 342}	{'precision': 0.9596774193548387, 'recall': 0.967479674796748, 'f1': 0.9635627530364373, 'number': 123}	{'precision': 0.9479166666666666, 'recall': 0.91, 'f1': 0.9285714285714285, 'number': 100}	{'precision': 0.990909090909091, 'recall': 0.8861788617886179, 'f1': 0.9356223175965666, 'number': 123}	{'precision': 1.0, 'recall': 0.8615384615384616, 'f1': 0.9256198347107438, 'number': 130}	{'precision': 0.987417943107221, 'recall': 0.9228016359918201, 'f1': 0.9540169133192389, 'number': 1956}	{'precision': 0.8166089965397924, 'recall': 0.9812889812889813, 'f1': 0.8914069877242682, 'number': 481}	{'precision': 0.09523809523809523, 'recall': 0.03571428571428571, 'f1': 0.051948051948051945, 'number': 56}	{'precision': 0.42907801418439717, 'recall': 0.952755905511811, 'f1': 0.5916870415647921, 'number': 127}	0.8989	0.9327	0.9155	0.9780
0.0399	10.0	190	0.0571	{'precision': 0.8933649289099526, 'recall': 0.9560439560439561, 'f1': 0.9236423029808084, 'number': 1183}	{'precision': 0.9659090909090909, 'recall': 0.9941520467836257, 'f1': 0.9798270893371758, 'number': 342}	{'precision': 0.9603174603174603, 'recall': 0.983739837398374, 'f1': 0.9718875502008032, 'number': 123}	{'precision': 0.94, 'recall': 0.94, 'f1': 0.94, 'number': 100}	{'precision': 1.0, 'recall': 0.9349593495934959, 'f1': 0.9663865546218486, 'number': 123}	{'precision': 1.0, 'recall': 0.9153846153846154, 'f1': 0.9558232931726908, 'number': 130}	{'precision': 0.9876277568585261, 'recall': 0.9386503067484663, 'f1': 0.9625163826998688, 'number': 1956}	{'precision': 0.8497316636851521, 'recall': 0.9875259875259875, 'f1': 0.9134615384615384, 'number': 481}	{'precision': 0.7368421052631579, 'recall': 0.5, 'f1': 0.5957446808510638, 'number': 56}	{'precision': 0.5757575757575758, 'recall': 0.8976377952755905, 'f1': 0.7015384615384614, 'number': 127}	0.9241	0.9463	0.9351	0.9831
0.0348	11.0	209	0.0652	{'precision': 0.8785549577248271, 'recall': 0.9661876584953508, 'f1': 0.9202898550724637, 'number': 1183}	{'precision': 0.9550561797752809, 'recall': 0.9941520467836257, 'f1': 0.9742120343839542, 'number': 342}	{'precision': 0.9453125, 'recall': 0.983739837398374, 'f1': 0.9641434262948206, 'number': 123}	{'precision': 0.94, 'recall': 0.94, 'f1': 0.94, 'number': 100}	{'precision': 0.9918032786885246, 'recall': 0.983739837398374, 'f1': 0.9877551020408164, 'number': 123}	{'precision': 0.984375, 'recall': 0.9692307692307692, 'f1': 0.9767441860465116, 'number': 130}	{'precision': 0.9912087912087912, 'recall': 0.9222903885480572, 'f1': 0.9555084745762713, 'number': 1956}	{'precision': 0.8191126279863481, 'recall': 0.997920997920998, 'f1': 0.8997188378631678, 'number': 481}	{'precision': 0.8367346938775511, 'recall': 0.7321428571428571, 'f1': 0.7809523809523811, 'number': 56}	{'precision': 0.711764705882353, 'recall': 0.952755905511811, 'f1': 0.8148148148148149, 'number': 127}	0.9225	0.9502	0.9361	0.9827
0.0302	12.0	228	0.0513	{'precision': 0.8866615265998458, 'recall': 0.9721048182586645, 'f1': 0.9274193548387096, 'number': 1183}	{'precision': 0.9798270893371758, 'recall': 0.9941520467836257, 'f1': 0.9869375907111755, 'number': 342}	{'precision': 0.9453125, 'recall': 0.983739837398374, 'f1': 0.9641434262948206, 'number': 123}	{'precision': 0.9393939393939394, 'recall': 0.93, 'f1': 0.9346733668341709, 'number': 100}	{'precision': 1.0, 'recall': 0.983739837398374, 'f1': 0.9918032786885246, 'number': 123}	{'precision': 0.9921875, 'recall': 0.9769230769230769, 'f1': 0.9844961240310077, 'number': 130}	{'precision': 0.9872, 'recall': 0.946319018404908, 'f1': 0.9663273296789351, 'number': 1956}	{'precision': 0.8574007220216606, 'recall': 0.9875259875259875, 'f1': 0.9178743961352657, 'number': 481}	{'precision': 0.8653846153846154, 'recall': 0.8035714285714286, 'f1': 0.8333333333333334, 'number': 56}	{'precision': 0.8345323741007195, 'recall': 0.9133858267716536, 'f1': 0.8721804511278195, 'number': 127}	0.9365	0.9606	0.9484	0.9861
0.0269	13.0	247	0.0543	{'precision': 0.9086576648133439, 'recall': 0.967032967032967, 'f1': 0.9369369369369369, 'number': 1183}	{'precision': 0.96045197740113, 'recall': 0.9941520467836257, 'f1': 0.9770114942528735, 'number': 342}	{'precision': 0.952755905511811, 'recall': 0.983739837398374, 'f1': 0.968, 'number': 123}	{'precision': 0.94, 'recall': 0.94, 'f1': 0.94, 'number': 100}	{'precision': 1.0, 'recall': 0.983739837398374, 'f1': 0.9918032786885246, 'number': 123}	{'precision': 0.9922480620155039, 'recall': 0.9846153846153847, 'f1': 0.9884169884169884, 'number': 130}	{'precision': 0.9866666666666667, 'recall': 0.9458077709611452, 'f1': 0.9658052727747326, 'number': 1956}	{'precision': 0.85431654676259, 'recall': 0.9875259875259875, 'f1': 0.9161041465766634, 'number': 481}	{'precision': 0.8545454545454545, 'recall': 0.8392857142857143, 'f1': 0.8468468468468467, 'number': 56}	{'precision': 0.8768115942028986, 'recall': 0.952755905511811, 'f1': 0.9132075471698112, 'number': 127}	0.9421	0.9610	0.9515	0.9872
0.0251	14.0	266	0.0549	{'precision': 0.8923315259488769, 'recall': 0.9737954353338969, 'f1': 0.931285367825384, 'number': 1183}	{'precision': 0.9855072463768116, 'recall': 0.9941520467836257, 'f1': 0.9898107714701602, 'number': 342}	{'precision': 0.9603174603174603, 'recall': 0.983739837398374, 'f1': 0.9718875502008032, 'number': 123}	{'precision': 0.9405940594059405, 'recall': 0.95, 'f1': 0.9452736318407959, 'number': 100}	{'precision': 1.0, 'recall': 0.983739837398374, 'f1': 0.9918032786885246, 'number': 123}	{'precision': 0.9922480620155039, 'recall': 0.9846153846153847, 'f1': 0.9884169884169884, 'number': 130}	{'precision': 0.98718633208756, 'recall': 0.9452965235173824, 'f1': 0.9657874118568818, 'number': 1956}	{'precision': 0.85431654676259, 'recall': 0.9875259875259875, 'f1': 0.9161041465766634, 'number': 481}	{'precision': 0.8421052631578947, 'recall': 0.8571428571428571, 'f1': 0.8495575221238938, 'number': 56}	{'precision': 0.7610062893081762, 'recall': 0.952755905511811, 'f1': 0.8461538461538461, 'number': 127}	0.9353	0.9630	0.9489	0.9862
0.0242	15.0	285	0.0543	{'precision': 0.902668759811617, 'recall': 0.9721048182586645, 'f1': 0.9361009361009361, 'number': 1183}	{'precision': 0.9855072463768116, 'recall': 0.9941520467836257, 'f1': 0.9898107714701602, 'number': 342}	{'precision': 0.9603174603174603, 'recall': 0.983739837398374, 'f1': 0.9718875502008032, 'number': 123}	{'precision': 0.9405940594059405, 'recall': 0.95, 'f1': 0.9452736318407959, 'number': 100}	{'precision': 1.0, 'recall': 0.983739837398374, 'f1': 0.9918032786885246, 'number': 123}	{'precision': 0.9922480620155039, 'recall': 0.9846153846153847, 'f1': 0.9884169884169884, 'number': 130}	{'precision': 0.9887580299785867, 'recall': 0.9442740286298569, 'f1': 0.9660041841004183, 'number': 1956}	{'precision': 0.8530465949820788, 'recall': 0.9896049896049897, 'f1': 0.9162656400384984, 'number': 481}	{'precision': 0.8421052631578947, 'recall': 0.8571428571428571, 'f1': 0.8495575221238938, 'number': 56}	{'precision': 0.7960526315789473, 'recall': 0.952755905511811, 'f1': 0.8673835125448028, 'number': 127}	0.9400	0.9623	0.9510	0.9870

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.0+cu121
• Datasets 3.0.0
• Tokenizers 0.19.1