lmv2-g-paystb-999-doc-09-11

This model is a fine-tuned version of microsoft/layoutlmv2-base-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1207
• Employee Address Precision: 0.8152
• Employee Address Recall: 0.8523
• Employee Address F1: 0.8333
• Employee Address Number: 88
• Employee Name Precision: 0.9511
• Employee Name Recall: 0.9722
• Employee Name F1: 0.9615
• Employee Name Number: 180
• Employer Address Precision: 0.8151
• Employer Address Recall: 0.8509
• Employer Address F1: 0.8326
• Employer Address Number: 114
• Employer Name Precision: 0.8564
• Employer Name Recall: 0.8857
• Employer Name F1: 0.8708
• Employer Name Number: 175
• Gross Pay Precision: 0.8976
• Gross Pay Recall: 0.8324
• Gross Pay F1: 0.8638
• Gross Pay Number: 179
• Net Pay Precision: 0.8994
• Net Pay Recall: 0.8846
• Net Pay F1: 0.8920
• Net Pay Number: 182
• Pay Date Precision: 0.9107
• Pay Date Recall: 0.8718
• Pay Date F1: 0.8908
• Pay Date Number: 117
• Ssn Number Precision: 0.9032
• Ssn Number Recall: 0.9032
• Ssn Number F1: 0.9032
• Ssn Number Number: 31
• Overall Precision: 0.8853
• Overall Recall: 0.8837
• Overall F1: 0.8845
• Overall Accuracy: 0.9834

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: 4e-05
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Employee Address Precision	Employee Address Recall	Employee Address F1	Employee Address Number	Employee Name Precision	Employee Name Recall	Employee Name F1	Employee Name Number	Employer Address Precision	Employer Address Recall	Employer Address F1	Employer Address Number	Employer Name Precision	Employer Name Recall	Employer Name F1	Employer Name Number	Gross Pay Precision	Gross Pay Recall	Gross Pay F1	Gross Pay Number	Net Pay Precision	Net Pay Recall	Net Pay F1	Net Pay Number	Pay Date Precision	Pay Date Recall	Pay Date F1	Pay Date Number	Ssn Number Precision	Ssn Number Recall	Ssn Number F1	Ssn Number Number	Overall Precision	Overall Recall	Overall F1	Overall Accuracy
0.8872	1.0	799	0.3053	0.0	0.0	0.0	88	0.5146	0.5889	0.5492	180	0.4213	0.7281	0.5338	114	0.6468	0.7429	0.6915	175	0.0	0.0	0.0	179	0.4143	0.4780	0.4439	182	0.0	0.0	0.0	117	0.8148	0.7097	0.7586	31	0.5089	0.4015	0.4489	0.9454
0.2088	2.0	1598	0.1402	0.7527	0.7955	0.7735	88	0.9270	0.9167	0.9218	180	0.7209	0.8158	0.7654	114	0.7949	0.8857	0.8378	175	0.8425	0.6872	0.7569	179	0.8079	0.7857	0.7967	182	0.7068	0.8034	0.752	117	0.8889	0.7742	0.8276	31	0.8043	0.8133	0.8088	0.9769
0.1044	3.0	2397	0.1181	0.6881	0.8523	0.7614	88	0.9227	0.9278	0.9252	180	0.7323	0.8158	0.7718	114	0.8297	0.8629	0.8459	175	0.8402	0.7933	0.8161	179	0.8132	0.8132	0.8132	182	0.8509	0.8291	0.8398	117	0.8333	0.8065	0.8197	31	0.8208	0.8424	0.8315	0.9780
0.0659	4.0	3196	0.1117	0.8105	0.875	0.8415	88	0.9101	0.9556	0.9322	180	0.784	0.8596	0.8201	114	0.8032	0.8629	0.8320	175	0.8247	0.8939	0.8579	179	0.7885	0.9011	0.8410	182	0.8120	0.9231	0.864	117	0.8966	0.8387	0.8667	31	0.8234	0.8968	0.8586	0.9791
0.0505	5.0	3995	0.0975	0.8636	0.8636	0.8636	88	0.9309	0.9722	0.9511	180	0.8115	0.8684	0.8390	114	0.8516	0.8857	0.8683	175	0.8772	0.8380	0.8571	179	0.8914	0.8571	0.8739	182	0.8333	0.9402	0.8835	117	0.8929	0.8065	0.8475	31	0.8711	0.8874	0.8792	0.9826
0.0407	6.0	4794	0.1120	0.7660	0.8182	0.7912	88	0.9301	0.9611	0.9454	180	0.7344	0.8246	0.7769	114	0.8245	0.8857	0.8540	175	0.8352	0.8492	0.8421	179	0.8261	0.8352	0.8306	182	0.8333	0.9402	0.8835	117	0.8788	0.9355	0.9062	31	0.8314	0.8790	0.8545	0.9794
0.032	7.0	5593	0.1585	0.7526	0.8295	0.7892	88	0.9454	0.9611	0.9532	180	0.8120	0.8333	0.8225	114	0.8457	0.8457	0.8457	175	0.8084	0.7542	0.7803	179	0.8426	0.5	0.6276	182	0.8271	0.9402	0.8800	117	0.8710	0.8710	0.8710	31	0.8427	0.7992	0.8204	0.9774
0.0298	8.0	6392	0.1335	0.7822	0.8977	0.8360	88	0.9344	0.95	0.9421	180	0.68	0.8947	0.7727	114	0.7536	0.8914	0.8168	175	0.8478	0.8715	0.8595	179	0.8876	0.8242	0.8547	182	0.8095	0.8718	0.8395	117	0.9	0.8710	0.8852	31	0.82	0.8846	0.8511	0.9765
0.0231	9.0	7191	0.1189	0.8315	0.8409	0.8362	88	0.9396	0.95	0.9448	180	0.7672	0.7807	0.7739	114	0.7989	0.84	0.8189	175	0.8508	0.8603	0.8556	179	0.8659	0.8516	0.8587	182	0.9217	0.9060	0.9138	117	0.9	0.8710	0.8852	31	0.8578	0.8659	0.8618	0.9814
0.0216	10.0	7990	0.1207	0.8152	0.8523	0.8333	88	0.9511	0.9722	0.9615	180	0.8151	0.8509	0.8326	114	0.8564	0.8857	0.8708	175	0.8976	0.8324	0.8638	179	0.8994	0.8846	0.8920	182	0.9107	0.8718	0.8908	117	0.9032	0.9032	0.9032	31	0.8853	0.8837	0.8845	0.9834
0.0229	11.0	8789	0.1297	0.8242	0.8523	0.8380	88	0.9663	0.9556	0.9609	180	0.8197	0.8772	0.8475	114	0.8772	0.8571	0.8671	175	0.8539	0.8492	0.8515	179	0.8817	0.8187	0.8490	182	0.9126	0.8034	0.8545	117	0.8056	0.9355	0.8657	31	0.8788	0.8640	0.8713	0.9816
0.0221	12.0	9588	0.1326	0.8462	0.875	0.8603	88	0.9318	0.9111	0.9213	180	0.7338	0.8947	0.8063	114	0.7487	0.8514	0.7968	175	0.8757	0.8659	0.8708	179	0.8871	0.9066	0.8967	182	0.8651	0.9316	0.8971	117	0.7436	0.9355	0.8286	31	0.8385	0.8912	0.8640	0.9810
0.0179	13.0	10387	0.1413	0.8085	0.8636	0.8352	88	0.9553	0.95	0.9526	180	0.8065	0.8772	0.8403	114	0.8098	0.8514	0.8301	175	0.8659	0.8659	0.8659	179	0.8729	0.8681	0.8705	182	0.8934	0.9316	0.9121	117	0.9333	0.9032	0.9180	31	0.8655	0.8874	0.8763	0.9825
0.0143	14.0	11186	0.1267	0.8315	0.8409	0.8362	88	0.9454	0.9611	0.9532	180	0.7372	0.8860	0.8048	114	0.8054	0.8514	0.8278	175	0.8043	0.8268	0.8154	179	0.7861	0.8681	0.8251	182	0.8889	0.8889	0.8889	117	0.8571	0.9677	0.9091	31	0.8285	0.8790	0.8530	0.9810
0.0139	15.0	11985	0.1592	0.7449	0.8295	0.7849	88	0.9355	0.9667	0.9508	180	0.8319	0.8246	0.8282	114	0.8125	0.8171	0.8148	175	0.8708	0.8659	0.8683	179	0.8944	0.8846	0.8895	182	0.9027	0.8718	0.8870	117	0.8788	0.9355	0.9062	31	0.8644	0.8734	0.8689	0.9807
0.0163	16.0	12784	0.1449	0.6496	0.8636	0.7415	88	0.8687	0.9556	0.9101	180	0.8448	0.8596	0.8522	114	0.7935	0.8343	0.8134	175	0.8168	0.8715	0.8432	179	0.8557	0.9121	0.8830	182	0.8413	0.9060	0.8724	117	0.8788	0.9355	0.9062	31	0.8188	0.8902	0.8530	0.9792
0.013	17.0	13583	0.1449	0.7353	0.8523	0.7895	88	0.9422	0.9056	0.9235	180	0.808	0.8860	0.8452	114	0.8207	0.8629	0.8412	175	0.8531	0.8436	0.8483	179	0.9416	0.7967	0.8631	182	0.9304	0.9145	0.9224	117	0.8571	0.9677	0.9091	31	0.8667	0.8659	0.8663	0.9805
0.0107	18.0	14382	0.1510	0.7315	0.8977	0.8061	88	0.9048	0.95	0.9268	180	0.8	0.8421	0.8205	114	0.8152	0.8571	0.8357	175	0.8844	0.8547	0.8693	179	0.8486	0.8626	0.8556	182	0.9076	0.9231	0.9153	117	0.8824	0.9677	0.9231	31	0.8489	0.8856	0.8669	0.9816
0.0108	19.0	15181	0.1424	0.7979	0.8523	0.8242	88	0.8895	0.9389	0.9135	180	0.8333	0.8772	0.8547	114	0.8075	0.8629	0.8343	175	0.8636	0.8492	0.8563	179	0.8596	0.8407	0.8500	182	0.9008	0.9316	0.9160	117	0.9032	0.9032	0.9032	31	0.8541	0.8790	0.8664	0.9829
0.0122	20.0	15980	0.1525	0.7228	0.8295	0.7725	88	0.9185	0.9389	0.9286	180	0.792	0.8684	0.8285	114	0.7513	0.8114	0.7802	175	0.8523	0.8380	0.8451	179	0.8870	0.8626	0.8747	182	0.8468	0.8974	0.8714	117	0.9655	0.9032	0.9333	31	0.8353	0.8659	0.8503	0.9813
0.0238	21.0	16779	0.1477	0.8605	0.8409	0.8506	88	0.9441	0.9389	0.9415	180	0.8279	0.8860	0.8559	114	0.8177	0.8457	0.8315	175	0.8523	0.8380	0.8451	179	0.9186	0.8681	0.8927	182	0.8417	0.8632	0.8523	117	0.8788	0.9355	0.9062	31	0.8700	0.8724	0.8712	0.9821
0.0124	22.0	17578	0.1458	0.7255	0.8409	0.7789	88	0.9435	0.9278	0.9356	180	0.7786	0.8947	0.8327	114	0.7968	0.8514	0.8232	175	0.8659	0.8659	0.8659	179	0.9029	0.8681	0.8852	182	0.8889	0.8889	0.8889	117	0.9032	0.9032	0.9032	31	0.8526	0.8790	0.8656	0.9819
0.0116	23.0	18377	0.1504	0.7895	0.8523	0.8197	88	0.9399	0.9556	0.9477	180	0.8065	0.8772	0.8403	114	0.7525	0.8686	0.8064	175	0.8655	0.8268	0.8457	179	0.8441	0.8626	0.8533	182	0.8	0.9231	0.8571	117	0.8333	0.9677	0.8955	31	0.8322	0.8837	0.8571	0.9803
0.0093	24.0	19176	0.1616	0.75	0.75	0.75	88	0.8978	0.9278	0.9126	180	0.8393	0.8246	0.8319	114	0.8261	0.8686	0.8468	175	0.8844	0.8547	0.8693	179	0.8757	0.8516	0.8635	182	0.8992	0.9145	0.9068	117	0.9062	0.9355	0.9206	31	0.8618	0.8659	0.8638	0.9804
0.0104	25.0	19975	0.1614	0.8182	0.8182	0.8182	88	0.9101	0.9556	0.9322	180	0.8197	0.8772	0.8475	114	0.8466	0.8514	0.8490	175	0.8475	0.8380	0.8427	179	0.8833	0.8736	0.8785	182	0.8548	0.9060	0.8797	117	0.8529	0.9355	0.8923	31	0.8596	0.8790	0.8692	0.9820
0.0105	26.0	20774	0.1362	0.7978	0.8068	0.8023	88	0.9076	0.9278	0.9176	180	0.8534	0.8684	0.8609	114	0.8541	0.9029	0.8778	175	0.8254	0.8715	0.8478	179	0.8852	0.8901	0.8877	182	0.9043	0.8889	0.8966	117	0.8485	0.9032	0.875	31	0.8638	0.8865	0.875	0.9822
0.0086	27.0	21573	0.1691	0.8172	0.8636	0.8398	88	0.9385	0.9333	0.9359	180	0.7407	0.8772	0.8032	114	0.7812	0.8571	0.8174	175	0.8539	0.8492	0.8515	179	0.875	0.8846	0.8798	182	0.9076	0.9231	0.9153	117	0.625	0.9677	0.7595	31	0.8378	0.8865	0.8614	0.9791
0.0092	28.0	22372	0.1536	0.7789	0.8409	0.8087	88	0.9266	0.9111	0.9188	180	0.8487	0.8860	0.8670	114	0.8588	0.8686	0.8636	175	0.8982	0.8380	0.8671	179	0.9	0.8901	0.8950	182	0.8783	0.8632	0.8707	117	0.9375	0.9677	0.9524	31	0.8795	0.8762	0.8778	0.9826
0.0065	29.0	23171	0.1676	0.8202	0.8295	0.8249	88	0.9444	0.9444	0.9444	180	0.7951	0.8509	0.8220	114	0.7685	0.8914	0.8254	175	0.9060	0.7542	0.8232	179	0.9153	0.8901	0.9025	182	0.9	0.8462	0.8722	117	0.9091	0.9677	0.9375	31	0.8674	0.8649	0.8661	0.9812
0.0118	30.0	23970	0.1803	0.8636	0.8636	0.8636	88	0.9293	0.95	0.9396	180	0.6690	0.8509	0.7490	114	0.8261	0.8686	0.8468	175	0.8	0.8715	0.8342	179	0.8421	0.8791	0.8602	182	0.8780	0.9231	0.9	117	0.8788	0.9355	0.9062	31	0.8310	0.8902	0.8596	0.9783

Framework versions

• Transformers 4.22.0.dev0
• Pytorch 1.12.1+cu113
• Datasets 2.2.2
• Tokenizers 0.12.1