lmv2-g-dl-243-doc-09-13

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: 0.2104
• Address Precision: 0.725
• Address Recall: 0.7632
• Address F1: 0.7436
• Address Number: 38
• Blood Group Precision: 0.8636
• Blood Group Recall: 0.8636
• Blood Group F1: 0.8636
• Blood Group Number: 22
• Date Of Issue Precision: 0.9787
• Date Of Issue Recall: 0.92
• Date Of Issue F1: 0.9485
• Date Of Issue Number: 50
• Dob Precision: 1.0
• Dob Recall: 0.9773
• Dob F1: 0.9885
• Dob Number: 44
• Driving Licence No Precision: 0.9796
• Driving Licence No Recall: 1.0
• Driving Licence No F1: 0.9897
• Driving Licence No Number: 48
• Name Precision: 0.9388
• Name Recall: 0.9388
• Name F1: 0.9388
• Name Number: 49
• S D W Name Precision: 0.9388
• S D W Name Recall: 0.9583
• S D W Name F1: 0.9485
• S D W Name Number: 48
• Valid Till Nt Precision: 0.7826
• Valid Till Nt Recall: 0.8780
• Valid Till Nt F1: 0.8276
• Valid Till Nt Number: 41
• Valid Till T Tr Precision: 0.9231
• Valid Till T Tr Recall: 0.8571
• Valid Till T Tr F1: 0.8889
• Valid Till T Tr Number: 14
• Overall Precision: 0.9078
• Overall Recall: 0.9181
• Overall F1: 0.9129
• Overall Accuracy: 0.9763

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	Address Precision	Address Recall	Address F1	Address Number	Blood Group Precision	Blood Group Recall	Blood Group F1	Blood Group Number	Date Of Issue Precision	Date Of Issue Recall	Date Of Issue F1	Date Of Issue Number	Dob Precision	Dob Recall	Dob F1	Dob Number	Driving Licence No Precision	Driving Licence No Recall	Driving Licence No F1	Driving Licence No Number	Name Precision	Name Recall	Name F1	Name Number	S D W Name Precision	S D W Name Recall	S D W Name F1	S D W Name Number	Valid Till Nt Precision	Valid Till Nt Recall	Valid Till Nt F1	Valid Till Nt Number	Valid Till T Tr Precision	Valid Till T Tr Recall	Valid Till T Tr F1	Valid Till T Tr Number	Overall Precision	Overall Recall	Overall F1	Overall Accuracy
1.9194	1.0	194	1.3597	0.3099	0.5789	0.4037	38	0.0	0.0	0.0	22	0.3147	0.9	0.4663	50	0.0	0.0	0.0	44	0.1059	0.1875	0.1353	48	0.0246	0.0612	0.0351	49	0.0	0.0	0.0	48	0.0	0.0	0.0	41	0.0	0.0	0.0	14	0.1876	0.2232	0.2039	0.8582
1.099	2.0	388	0.7767	0.5333	0.6316	0.5783	38	0.0	0.0	0.0	22	0.4667	0.98	0.6323	50	0.9773	0.9773	0.9773	44	0.9057	1.0	0.9505	48	0.3070	0.7143	0.4294	49	0.0	0.0	0.0	48	0.8182	0.2195	0.3462	41	0.0	0.0	0.0	14	0.5591	0.5876	0.5730	0.9016
0.6398	3.0	582	0.4892	0.5532	0.6842	0.6118	38	0.0	0.0	0.0	22	1.0	0.94	0.9691	50	1.0	0.9091	0.9524	44	0.9412	1.0	0.9697	48	0.5538	0.7347	0.6316	49	0.5714	0.75	0.6486	48	0.6863	0.8537	0.7609	41	0.0	0.0	0.0	14	0.7363	0.7571	0.7465	0.9491
0.4251	4.0	776	0.3770	0.4364	0.6316	0.5161	38	0.4706	0.3636	0.4103	22	1.0	0.9	0.9474	50	1.0	0.9773	0.9885	44	0.9412	1.0	0.9697	48	0.92	0.9388	0.9293	49	0.9333	0.875	0.9032	48	0.6182	0.8293	0.7083	41	0.0	0.0	0.0	14	0.8033	0.8192	0.8112	0.9557
0.304	5.0	970	0.2942	0.65	0.6842	0.6667	38	0.4643	0.5909	0.52	22	0.9796	0.96	0.9697	50	1.0	0.9773	0.9885	44	0.8727	1.0	0.9320	48	0.9388	0.9388	0.9388	49	0.9388	0.9583	0.9485	48	0.85	0.8293	0.8395	41	0.8	0.2857	0.4211	14	0.8603	0.8701	0.8652	0.9627
0.2297	6.0	1164	0.2497	0.5854	0.6316	0.6076	38	0.5294	0.8182	0.6429	22	1.0	0.92	0.9583	50	1.0	0.9773	0.9885	44	0.9231	1.0	0.9600	48	0.9149	0.8776	0.8958	49	0.9184	0.9375	0.9278	48	0.8462	0.8049	0.8250	41	0.7692	0.7143	0.7407	14	0.8516	0.8757	0.8635	0.9679
0.1843	7.0	1358	0.2123	0.54	0.7105	0.6136	38	0.7778	0.9545	0.8571	22	0.9792	0.94	0.9592	50	1.0	0.9773	0.9885	44	0.9796	1.0	0.9897	48	0.9348	0.8776	0.9053	49	0.9020	0.9583	0.9293	48	0.7609	0.8537	0.8046	41	0.8	0.5714	0.6667	14	0.8595	0.8983	0.8785	0.9696
0.1455	8.0	1552	0.2166	0.6316	0.6316	0.6316	38	0.8077	0.9545	0.875	22	1.0	0.94	0.9691	50	1.0	0.9773	0.9885	44	0.9216	0.9792	0.9495	48	0.9167	0.8980	0.9072	49	0.9167	0.9167	0.9167	48	0.7826	0.8780	0.8276	41	0.9286	0.9286	0.9286	14	0.8837	0.9011	0.8923	0.9696
0.1262	9.0	1746	0.2060	0.5333	0.6316	0.5783	38	0.8333	0.9091	0.8696	22	0.96	0.96	0.96	50	1.0	0.9773	0.9885	44	0.9216	0.9792	0.9495	48	0.9	0.9184	0.9091	49	0.9184	0.9375	0.9278	48	0.7826	0.8780	0.8276	41	0.5185	1.0	0.6829	14	0.8364	0.9096	0.8714	0.9661
0.109	10.0	1940	0.1912	0.6154	0.6316	0.6234	38	0.7692	0.9091	0.8333	22	0.9792	0.94	0.9592	50	1.0	0.9773	0.9885	44	0.9796	1.0	0.9897	48	0.9388	0.9388	0.9388	49	0.9362	0.9167	0.9263	48	0.8537	0.8537	0.8537	41	0.9333	1.0	0.9655	14	0.8992	0.9068	0.9030	0.9725
0.0911	11.0	2134	0.2063	0.5897	0.6053	0.5974	38	0.8	0.9091	0.8511	22	0.9412	0.96	0.9505	50	1.0	0.9773	0.9885	44	0.9412	1.0	0.9697	48	0.9184	0.9184	0.9184	49	0.9375	0.9375	0.9375	48	0.8	0.8780	0.8372	41	0.8235	1.0	0.9032	14	0.875	0.9096	0.8920	0.9690
0.0771	12.0	2328	0.2262	0.525	0.5526	0.5385	38	0.8333	0.9091	0.8696	22	1.0	0.98	0.9899	50	1.0	0.9773	0.9885	44	0.9184	0.9375	0.9278	48	0.9388	0.9388	0.9388	49	0.9375	0.9375	0.9375	48	0.8537	0.8537	0.8537	41	0.8571	0.8571	0.8571	14	0.8852	0.8927	0.8889	0.9638
0.0753	13.0	2522	0.2170	0.5714	0.6316	0.6	38	0.8	0.9091	0.8511	22	0.9796	0.96	0.9697	50	1.0	0.9773	0.9885	44	0.9412	1.0	0.9697	48	0.9375	0.9184	0.9278	49	0.9574	0.9375	0.9474	48	0.875	0.8537	0.8642	41	0.7059	0.8571	0.7742	14	0.8840	0.9040	0.8939	0.9673
0.0676	14.0	2716	0.2148	0.5610	0.6053	0.5823	38	0.8261	0.8636	0.8444	22	0.9245	0.98	0.9515	50	1.0	0.9773	0.9885	44	0.96	1.0	0.9796	48	0.9362	0.8980	0.9167	49	0.9130	0.875	0.8936	48	0.8182	0.8780	0.8471	41	0.9333	1.0	0.9655	14	0.8785	0.8983	0.8883	0.9670
0.0588	15.0	2910	0.2140	0.65	0.6842	0.6667	38	0.7241	0.9545	0.8235	22	0.9792	0.94	0.9592	50	1.0	0.9773	0.9885	44	0.9412	1.0	0.9697	48	0.8824	0.9184	0.9	49	0.9130	0.875	0.8936	48	0.8	0.8780	0.8372	41	0.9286	0.9286	0.9286	14	0.8747	0.9068	0.8904	0.9690
0.0592	16.0	3104	0.2353	0.6410	0.6579	0.6494	38	0.75	0.9545	0.84	22	0.9767	0.84	0.9032	50	1.0	0.9773	0.9885	44	0.9796	1.0	0.9897	48	0.9362	0.8980	0.9167	49	0.9574	0.9375	0.9474	48	0.8571	0.8780	0.8675	41	0.9333	1.0	0.9655	14	0.9008	0.8983	0.8996	0.9664
0.0461	17.0	3298	0.2137	0.5714	0.6316	0.6	38	0.7143	0.9091	0.8	22	0.9057	0.96	0.9320	50	1.0	0.9773	0.9885	44	0.9796	1.0	0.9897	48	0.9388	0.9388	0.9388	49	0.9184	0.9375	0.9278	48	0.8571	0.8780	0.8675	41	0.7368	1.0	0.8485	14	0.8663	0.9153	0.8901	0.9685
0.0424	18.0	3492	0.2057	0.5610	0.6053	0.5823	38	0.84	0.9545	0.8936	22	0.9423	0.98	0.9608	50	1.0	0.9773	0.9885	44	0.9796	1.0	0.9897	48	0.9388	0.9388	0.9388	49	0.9565	0.9167	0.9362	48	0.8095	0.8293	0.8193	41	0.8667	0.9286	0.8966	14	0.8867	0.9068	0.8966	0.9708
0.0389	19.0	3686	0.2400	0.6098	0.6579	0.6329	38	0.8	0.9091	0.8511	22	0.9574	0.9	0.9278	50	1.0	0.9773	0.9885	44	0.9796	1.0	0.9897	48	0.86	0.8776	0.8687	49	0.8980	0.9167	0.9072	48	0.8537	0.8537	0.8537	41	0.9167	0.7857	0.8462	14	0.8796	0.8870	0.8833	0.9670
0.0375	20.0	3880	0.2258	0.6341	0.6842	0.6582	38	0.8636	0.8636	0.8636	22	1.0	0.96	0.9796	50	1.0	0.9773	0.9885	44	0.9796	1.0	0.9897	48	0.9388	0.9388	0.9388	49	0.9565	0.9167	0.9362	48	0.8718	0.8293	0.8500	41	0.75	0.8571	0.8000	14	0.9065	0.9040	0.9052	0.9693
0.036	21.0	4074	0.2686	0.5952	0.6579	0.625	38	0.7778	0.9545	0.8571	22	1.0	0.96	0.9796	50	1.0	0.9773	0.9885	44	0.9796	1.0	0.9897	48	0.9362	0.8980	0.9167	49	0.9565	0.9167	0.9362	48	0.8684	0.8049	0.8354	41	0.7059	0.8571	0.7742	14	0.8908	0.8983	0.8945	0.9644
0.0321	22.0	4268	0.2102	0.6923	0.7105	0.7013	38	0.7778	0.9545	0.8571	22	0.96	0.96	0.96	50	1.0	0.9773	0.9885	44	0.9796	1.0	0.9897	48	0.92	0.9388	0.9293	49	0.9362	0.9167	0.9263	48	0.8182	0.8780	0.8471	41	0.8571	0.8571	0.8571	14	0.8953	0.9181	0.9066	0.9722
0.0244	23.0	4462	0.2432	0.6279	0.7105	0.6667	38	0.8	0.9091	0.8511	22	1.0	0.86	0.9247	50	1.0	0.9773	0.9885	44	0.9796	1.0	0.9897	48	0.9388	0.9388	0.9388	49	0.9362	0.9167	0.9263	48	0.7955	0.8537	0.8235	41	0.9091	0.7143	0.8	14	0.8927	0.8927	0.8927	0.9676
0.0217	24.0	4656	0.2290	0.6923	0.7105	0.7013	38	0.8261	0.8636	0.8444	22	1.0	0.92	0.9583	50	1.0	0.9773	0.9885	44	0.9796	1.0	0.9897	48	0.9388	0.9388	0.9388	49	0.9375	0.9375	0.9375	48	0.8333	0.8537	0.8434	41	0.9167	0.7857	0.8462	14	0.9117	0.9040	0.9078	0.9728
0.0215	25.0	4850	0.2677	0.5897	0.6053	0.5974	38	0.7778	0.9545	0.8571	22	0.9592	0.94	0.9495	50	1.0	0.9773	0.9885	44	0.9796	1.0	0.9897	48	0.9388	0.9388	0.9388	49	0.9375	0.9375	0.9375	48	0.7955	0.8537	0.8235	41	0.7	1.0	0.8235	14	0.875	0.9096	0.8920	0.9676
0.0258	26.0	5044	0.2356	0.6341	0.6842	0.6582	38	0.7692	0.9091	0.8333	22	0.9796	0.96	0.9697	50	1.0	0.9773	0.9885	44	0.9796	1.0	0.9897	48	0.9388	0.9388	0.9388	49	0.8980	0.9167	0.9072	48	0.7727	0.8293	0.8000	41	0.6923	0.6429	0.6667	14	0.8760	0.8983	0.8870	0.9696
0.0191	27.0	5238	0.2115	0.625	0.6579	0.6410	38	0.8696	0.9091	0.8889	22	0.9792	0.94	0.9592	50	1.0	0.9773	0.9885	44	0.9796	1.0	0.9897	48	0.9388	0.9388	0.9388	49	0.9375	0.9375	0.9375	48	0.8571	0.8780	0.8675	41	0.8	0.8571	0.8276	14	0.9020	0.9096	0.9058	0.9751
0.0238	28.0	5432	0.2104	0.725	0.7632	0.7436	38	0.8636	0.8636	0.8636	22	0.9787	0.92	0.9485	50	1.0	0.9773	0.9885	44	0.9796	1.0	0.9897	48	0.9388	0.9388	0.9388	49	0.9388	0.9583	0.9485	48	0.7826	0.8780	0.8276	41	0.9231	0.8571	0.8889	14	0.9078	0.9181	0.9129	0.9763
0.0282	29.0	5626	0.2352	0.5238	0.5789	0.5500	38	0.8696	0.9091	0.8889	22	0.9778	0.88	0.9263	50	1.0	0.9773	0.9885	44	0.9796	1.0	0.9897	48	0.9388	0.9388	0.9388	49	0.9362	0.9167	0.9263	48	0.7609	0.8537	0.8046	41	0.75	0.8571	0.8000	14	0.8722	0.8870	0.8796	0.9705
0.0157	30.0	5820	0.2614	0.6190	0.6842	0.6500	38	0.84	0.9545	0.8936	22	1.0	0.92	0.9583	50	1.0	0.9773	0.9885	44	0.9796	1.0	0.9897	48	0.9388	0.9388	0.9388	49	0.9583	0.9583	0.9583	48	0.72	0.8780	0.7912	41	0.7059	0.8571	0.7742	14	0.8780	0.9153	0.8963	0.9688

Framework versions

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