LayoutLMv3_1

This model is a fine-tuned version of microsoft/layoutlmv3-large on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3772
• Precision: 0.7355
• Recall: 0.7550
• F1: 0.7451
• Accuracy: 0.9035

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: 1e-06
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 1500

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	0.37	10	2.5352	0.0089	0.0199	0.0123	0.0503
No log	0.74	20	2.1493	0.0377	0.0397	0.0387	0.6028
No log	1.11	30	1.7234	0.0	0.0	0.0	0.7804
No log	1.48	40	1.2971	0.0	0.0	0.0	0.7818
No log	1.85	50	1.0544	0.0	0.0	0.0	0.7818
No log	2.22	60	1.0210	0.0	0.0	0.0	0.7818
No log	2.59	70	0.9842	0.0	0.0	0.0	0.7818
No log	2.96	80	0.9651	0.0	0.0	0.0	0.7818
No log	3.33	90	0.9402	0.0	0.0	0.0	0.7818
No log	3.7	100	0.9205	0.0	0.0	0.0	0.7818
No log	4.07	110	0.9035	0.0	0.0	0.0	0.7818
No log	4.44	120	0.8807	0.0	0.0	0.0	0.7818
No log	4.81	130	0.8596	0.0	0.0	0.0	0.7818
No log	5.19	140	0.8382	0.0	0.0	0.0	0.7818
No log	5.56	150	0.8151	0.0	0.0	0.0	0.7818
No log	5.93	160	0.8039	0.0	0.0	0.0	0.7818
No log	6.3	170	0.7864	0.0	0.0	0.0	0.7818
No log	6.67	180	0.7588	0.0	0.0	0.0	0.7818
No log	7.04	190	0.7352	0.0	0.0	0.0	0.7818
No log	7.41	200	0.7232	0.0	0.0	0.0	0.7818
No log	7.78	210	0.7246	1.0	0.0132	0.0261	0.7846
No log	8.15	220	0.6916	1.0	0.0132	0.0261	0.7846
No log	8.52	230	0.6755	1.0	0.0199	0.0390	0.7860
No log	8.89	240	0.6686	1.0	0.0331	0.0641	0.7888
No log	9.26	250	0.6514	1.0	0.0265	0.0516	0.7874
No log	9.63	260	0.6430	1.0	0.0596	0.1125	0.7944
No log	10.0	270	0.6280	0.9474	0.1192	0.2118	0.8070
No log	10.37	280	0.6127	0.9286	0.1722	0.2905	0.8196
No log	10.74	290	0.6136	0.9024	0.2450	0.3854	0.8350
No log	11.11	300	0.5886	0.8810	0.2450	0.3834	0.8350
No log	11.48	310	0.5896	0.8909	0.3245	0.4757	0.8517
No log	11.85	320	0.5732	0.9310	0.3576	0.5167	0.8587
No log	12.22	330	0.5770	0.8533	0.4238	0.5664	0.8671
No log	12.59	340	0.5557	0.8649	0.4238	0.5689	0.8671
No log	12.96	350	0.5469	0.8222	0.4901	0.6141	0.8797
No log	13.33	360	0.5412	0.8242	0.4967	0.6198	0.8825
No log	13.7	370	0.5313	0.8454	0.5430	0.6613	0.8923
No log	14.07	380	0.5188	0.8381	0.5828	0.6875	0.8993
No log	14.44	390	0.5190	0.8333	0.5960	0.6950	0.8993
No log	14.81	400	0.5172	0.8165	0.5894	0.6846	0.8993
No log	15.19	410	0.5066	0.7966	0.6225	0.6989	0.9021
No log	15.56	420	0.4879	0.8087	0.6159	0.6992	0.9035
No log	15.93	430	0.4943	0.7833	0.6225	0.6937	0.9021
No log	16.3	440	0.4716	0.8205	0.6358	0.7164	0.9091
No log	16.67	450	0.4594	0.8264	0.6623	0.7353	0.9119
No log	17.04	460	0.4758	0.7761	0.6887	0.7298	0.9105
No log	17.41	470	0.4520	0.8430	0.6755	0.75	0.9217
No log	17.78	480	0.4582	0.8244	0.7152	0.7660	0.9245
No log	18.15	490	0.4475	0.8189	0.6887	0.7482	0.9217
0.6982	18.52	500	0.4627	0.7431	0.7086	0.7254	0.9105
0.6982	18.89	510	0.4419	0.7826	0.7152	0.7474	0.9189
0.6982	19.26	520	0.4351	0.7730	0.7219	0.7466	0.9147
0.6982	19.63	530	0.4213	0.7857	0.7285	0.7560	0.9189
0.6982	20.0	540	0.4389	0.7273	0.7417	0.7344	0.9091
0.6982	20.37	550	0.4208	0.7762	0.7351	0.7551	0.9189
0.6982	20.74	560	0.4301	0.74	0.7351	0.7375	0.9119
0.6982	21.11	570	0.4199	0.7568	0.7417	0.7492	0.9161
0.6982	21.48	580	0.4283	0.7006	0.7285	0.7143	0.9021
0.6982	21.85	590	0.4068	0.7857	0.7285	0.7560	0.9203
0.6982	22.22	600	0.4241	0.7179	0.7417	0.7296	0.9077
0.6982	22.59	610	0.3988	0.8321	0.7550	0.7917	0.9329
0.6982	22.96	620	0.4005	0.7671	0.7417	0.7542	0.9189
0.6982	23.33	630	0.3939	0.7651	0.7550	0.76	0.9189
0.6982	23.7	640	0.4007	0.7278	0.7616	0.7443	0.9119
0.6982	24.07	650	0.3857	0.7973	0.7815	0.7893	0.9217
0.6982	24.44	660	0.3893	0.7682	0.7682	0.7682	0.9175
0.6982	24.81	670	0.3946	0.7516	0.7616	0.7566	0.9147
0.6982	25.19	680	0.3893	0.7516	0.7616	0.7566	0.9161
0.6982	25.56	690	0.3969	0.7419	0.7616	0.7516	0.9119
0.6982	25.93	700	0.3854	0.7852	0.7748	0.7800	0.9217
0.6982	26.3	710	0.3858	0.7973	0.7815	0.7893	0.9231
0.6982	26.67	720	0.3831	0.7867	0.7815	0.7841	0.9217
0.6982	27.04	730	0.3996	0.7267	0.7748	0.75	0.9049
0.6982	27.41	740	0.3907	0.7358	0.7748	0.7548	0.9077
0.6982	27.78	750	0.3720	0.8013	0.8013	0.8013	0.9245
0.6982	28.15	760	0.3799	0.7895	0.7947	0.7921	0.9189
0.6982	28.52	770	0.3938	0.7178	0.7748	0.7452	0.9035
0.6982	28.89	780	0.3761	0.7763	0.7815	0.7789	0.9189
0.6982	29.26	790	0.3906	0.7267	0.7748	0.75	0.9063
0.6982	29.63	800	0.3780	0.7436	0.7682	0.7557	0.9105
0.6982	30.0	810	0.3773	0.7548	0.7748	0.7647	0.9133
0.6982	30.37	820	0.3716	0.7727	0.7881	0.7803	0.9175
0.6982	30.74	830	0.3747	0.7452	0.7748	0.7597	0.9119
0.6982	31.11	840	0.3747	0.7405	0.7748	0.7573	0.9133
0.6982	31.48	850	0.3821	0.7239	0.7815	0.7516	0.9077
0.6982	31.85	860	0.3649	0.7697	0.7748	0.7723	0.9175
0.6982	32.22	870	0.3804	0.7152	0.7815	0.7468	0.9049
0.6982	32.59	880	0.3715	0.75	0.7748	0.7622	0.9105
0.6982	32.96	890	0.3663	0.7632	0.7682	0.7657	0.9161
0.6982	33.33	900	0.3713	0.7516	0.7815	0.7662	0.9133
0.6982	33.7	910	0.3684	0.7597	0.7748	0.7672	0.9133
0.6982	34.07	920	0.3708	0.75	0.7748	0.7622	0.9119
0.6982	34.44	930	0.3699	0.8146	0.8146	0.8146	0.9259
0.6982	34.81	940	0.3726	0.7778	0.7881	0.7829	0.9189
0.6982	35.19	950	0.3763	0.7405	0.7748	0.7573	0.9105
0.6982	35.56	960	0.3883	0.7267	0.7748	0.75	0.9035
0.6982	35.93	970	0.3729	0.7616	0.7616	0.7616	0.9119
0.6982	36.3	980	0.3654	0.8108	0.7947	0.8027	0.9217
0.6982	36.67	990	0.3795	0.7195	0.7815	0.7492	0.9049
0.2195	37.04	1000	0.3819	0.7267	0.7748	0.75	0.9035
0.2195	37.41	1010	0.3760	0.7233	0.7616	0.7419	0.9035
0.2195	37.78	1020	0.3664	0.7468	0.7616	0.7541	0.9105
0.2195	38.15	1030	0.3753	0.7312	0.7748	0.7524	0.9077
0.2195	38.52	1040	0.3791	0.7284	0.7815	0.7540	0.9035
0.2195	38.89	1050	0.3665	0.7933	0.7881	0.7907	0.9203
0.2195	39.26	1060	0.3655	0.7763	0.7815	0.7789	0.9161
0.2195	39.63	1070	0.3811	0.7312	0.7748	0.7524	0.9091
0.2195	40.0	1080	0.3725	0.7342	0.7682	0.7508	0.9063
0.2195	40.37	1090	0.3639	0.7692	0.7947	0.7818	0.9161
0.2195	40.74	1100	0.3721	0.7312	0.7748	0.7524	0.9077
0.2195	41.11	1110	0.3782	0.7143	0.7616	0.7372	0.9021
0.2195	41.48	1120	0.3654	0.7748	0.7748	0.7748	0.9189
0.2195	41.85	1130	0.3717	0.7278	0.7616	0.7443	0.9049
0.2195	42.22	1140	0.3868	0.7195	0.7815	0.7492	0.9021
0.2195	42.59	1150	0.3913	0.7066	0.7815	0.7421	0.8993
0.2195	42.96	1160	0.3797	0.7222	0.7748	0.7476	0.9035
0.2195	43.33	1170	0.3709	0.7405	0.7748	0.7573	0.9105
0.2195	43.7	1180	0.3736	0.7358	0.7748	0.7548	0.9077
0.2195	44.07	1190	0.3664	0.7389	0.7682	0.7532	0.9091
0.2195	44.44	1200	0.3677	0.7358	0.7748	0.7548	0.9063
0.2195	44.81	1210	0.3805	0.7329	0.7815	0.7564	0.9077
0.2195	45.19	1220	0.3806	0.7329	0.7815	0.7564	0.9077
0.2195	45.56	1230	0.3712	0.7372	0.7616	0.7492	0.9035
0.2195	45.93	1240	0.3746	0.7308	0.7550	0.7427	0.9035
0.2195	46.3	1250	0.3725	0.7261	0.7550	0.7403	0.9049
0.2195	46.67	1260	0.3719	0.7355	0.7550	0.7451	0.9035
0.2195	47.04	1270	0.3718	0.7355	0.7550	0.7451	0.9063
0.2195	47.41	1280	0.3728	0.7355	0.7550	0.7451	0.9063
0.2195	47.78	1290	0.3740	0.7261	0.7550	0.7403	0.9035
0.2195	48.15	1300	0.3780	0.7325	0.7616	0.7468	0.9035
0.2195	48.52	1310	0.3796	0.7325	0.7616	0.7468	0.9035
0.2195	48.89	1320	0.3816	0.7325	0.7616	0.7468	0.9035
0.2195	49.26	1330	0.3816	0.7278	0.7616	0.7443	0.9021
0.2195	49.63	1340	0.3803	0.7278	0.7616	0.7443	0.9021
0.2195	50.0	1350	0.3777	0.7308	0.7550	0.7427	0.9049
0.2195	50.37	1360	0.3810	0.7325	0.7616	0.7468	0.9035
0.2195	50.74	1370	0.3793	0.7325	0.7616	0.7468	0.9063
0.2195	51.11	1380	0.3773	0.7308	0.7550	0.7427	0.9049
0.2195	51.48	1390	0.3791	0.7342	0.7682	0.7508	0.9049
0.2195	51.85	1400	0.3822	0.7342	0.7682	0.7508	0.9049
0.2195	52.22	1410	0.3830	0.7342	0.7682	0.7508	0.9049
0.2195	52.59	1420	0.3797	0.7342	0.7682	0.7508	0.9049
0.2195	52.96	1430	0.3791	0.7342	0.7682	0.7508	0.9049
0.2195	53.33	1440	0.3790	0.7342	0.7682	0.7508	0.9049
0.2195	53.7	1450	0.3792	0.7342	0.7682	0.7508	0.9049
0.2195	54.07	1460	0.3786	0.7325	0.7616	0.7468	0.9035
0.2195	54.44	1470	0.3778	0.7355	0.7550	0.7451	0.9035
0.2195	54.81	1480	0.3776	0.7355	0.7550	0.7451	0.9035
0.2195	55.19	1490	0.3774	0.7355	0.7550	0.7451	0.9035
0.1305	55.56	1500	0.3772	0.7355	0.7550	0.7451	0.9035

Framework versions

• Transformers 4.29.2
• Pytorch 2.0.1+cu118
• Datasets 2.14.4
• Tokenizers 0.13.3