model10.0

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

• Loss: 0.3117
• Precision: 0.7824
• Recall: 0.7396
• F1: 0.7604
• Accuracy: 0.9481

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-05
• 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: 10000

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	0.1094	100	1.1070	0.0820	0.0037	0.0072	0.8202
No log	0.2188	200	0.9957	0.0449	0.0049	0.0088	0.8244
No log	0.3282	300	0.9355	0.1949	0.0590	0.0906	0.8338
No log	0.4376	400	0.8255	0.2373	0.1223	0.1614	0.8453
1.0172	0.5470	500	0.7511	0.3185	0.1510	0.2049	0.8489
1.0172	0.6565	600	0.7203	0.3609	0.1997	0.2571	0.8510
1.0172	0.7659	700	0.6571	0.4256	0.2565	0.3200	0.8755
1.0172	0.8753	800	0.5845	0.4510	0.3127	0.3693	0.8873
1.0172	0.9847	900	0.5614	0.5053	0.3477	0.4119	0.8948
0.45	1.0941	1000	0.5126	0.5561	0.3818	0.4527	0.9042
0.45	1.2035	1100	0.5350	0.5790	0.4176	0.4852	0.9092
0.45	1.3129	1200	0.4846	0.5449	0.4562	0.4966	0.9068
0.45	1.4223	1300	0.4681	0.6153	0.4843	0.5420	0.9149
0.45	1.5317	1400	0.4807	0.6572	0.4874	0.5598	0.9164
0.2388	1.6411	1500	0.4096	0.5848	0.5264	0.5541	0.9146
0.2388	1.7505	1600	0.3991	0.6344	0.5474	0.5877	0.9203
0.2388	1.8600	1700	0.4044	0.5996	0.5691	0.5840	0.9185
0.2388	1.9694	1800	0.4065	0.6607	0.5869	0.6216	0.9258
0.2388	2.0788	1900	0.4159	0.6018	0.5802	0.5908	0.9123
0.1768	2.1882	2000	0.4203	0.6846	0.5822	0.6293	0.9268
0.1768	2.2976	2100	0.3934	0.6566	0.5954	0.6245	0.9276
0.1768	2.4070	2200	0.3879	0.7137	0.6038	0.6542	0.9313
0.1768	2.5164	2300	0.3829	0.5685	0.6326	0.5989	0.9175
0.1768	2.6258	2400	0.3508	0.7175	0.6191	0.6647	0.9328
0.1399	2.7352	2500	0.3215	0.6869	0.6311	0.6578	0.9327
0.1399	2.8446	2600	0.3271	0.7248	0.6171	0.6666	0.9358
0.1399	2.9540	2700	0.3226	0.6491	0.6544	0.6517	0.9277
0.1399	3.0635	2800	0.3336	0.6596	0.6386	0.6490	0.9278
0.1399	3.1729	2900	0.3423	0.6480	0.6624	0.6551	0.9314
0.1083	3.2823	3000	0.3698	0.7509	0.6566	0.7006	0.9372
0.1083	3.3917	3100	0.3353	0.6457	0.6649	0.6552	0.9287
0.1083	3.5011	3200	0.3391	0.7518	0.6626	0.7044	0.9383
0.1083	3.6105	3300	0.3314	0.7350	0.6699	0.7010	0.9381
0.1083	3.7199	3400	0.3338	0.6728	0.6832	0.6779	0.9347
0.0988	3.8293	3500	0.3239	0.7509	0.6753	0.7111	0.9369
0.0988	3.9387	3600	0.3481	0.7555	0.6564	0.7025	0.9395
0.0988	4.0481	3700	0.3231	0.6749	0.6883	0.6815	0.9348
0.0988	4.1575	3800	0.3581	0.7669	0.6699	0.7151	0.9411
0.0988	4.2670	3900	0.3213	0.7174	0.6873	0.7021	0.9389
0.0775	4.3764	4000	0.3244	0.7433	0.6738	0.7069	0.9387
0.0775	4.4858	4100	0.3275	0.7370	0.6868	0.7110	0.9405
0.0775	4.5952	4200	0.3197	0.7405	0.6997	0.7195	0.9413
0.0775	4.7046	4300	0.3183	0.7419	0.6935	0.7169	0.9415
0.0775	4.8140	4400	0.2961	0.7445	0.6933	0.7180	0.9408
0.0771	4.9234	4500	0.3195	0.7542	0.6986	0.7253	0.9426
0.0771	5.0328	4600	0.3295	0.7637	0.7010	0.7310	0.9435
0.0771	5.1422	4700	0.3204	0.7603	0.7006	0.7293	0.9434
0.0771	5.2516	4800	0.2992	0.7443	0.6995	0.7212	0.9395
0.0771	5.3611	4900	0.2978	0.7312	0.7033	0.7170	0.9393
0.0647	5.4705	5000	0.3324	0.7608	0.7079	0.7334	0.9432
0.0647	5.5799	5100	0.3356	0.7635	0.7038	0.7324	0.9430
0.0647	5.6893	5200	0.3121	0.7634	0.7121	0.7368	0.9430
0.0647	5.7987	5300	0.3392	0.7858	0.7003	0.7406	0.9448
0.0647	5.9081	5400	0.2952	0.7265	0.7220	0.7242	0.9412
0.0573	6.0175	5500	0.3070	0.7311	0.7211	0.7260	0.9429
0.0573	6.1269	5600	0.3207	0.7414	0.7241	0.7326	0.9435
0.0573	6.2363	5700	0.3130	0.7685	0.7231	0.7451	0.9455
0.0573	6.3457	5800	0.3441	0.7752	0.7139	0.7433	0.9447
0.0573	6.4551	5900	0.3196	0.7818	0.7128	0.7457	0.9458
0.0529	6.5646	6000	0.3369	0.7907	0.7164	0.7517	0.9456
0.0529	6.6740	6100	0.3059	0.7394	0.7267	0.7330	0.9435
0.0529	6.7834	6200	0.3043	0.7624	0.7231	0.7422	0.9444
0.0529	6.8928	6300	0.3028	0.7527	0.7252	0.7387	0.9441
0.0529	7.0022	6400	0.3089	0.7596	0.7293	0.7441	0.9457
0.0542	7.1116	6500	0.2927	0.7306	0.7286	0.7296	0.9408
0.0542	7.2210	6600	0.3178	0.7785	0.7274	0.7521	0.9456
0.0542	7.3304	6700	0.3267	0.7653	0.7304	0.7474	0.9450
0.0542	7.4398	6800	0.3254	0.7618	0.7280	0.7445	0.9450
0.0542	7.5492	6900	0.3240	0.7856	0.7254	0.7543	0.9464
0.0416	7.6586	7000	0.3203	0.7682	0.7319	0.7496	0.9463
0.0416	7.7681	7100	0.3176	0.7801	0.7299	0.7542	0.9468
0.0416	7.8775	7200	0.3012	0.7601	0.7355	0.7476	0.9470
0.0416	7.9869	7300	0.3092	0.7336	0.7377	0.7357	0.9436
0.0416	8.0963	7400	0.3025	0.7782	0.7349	0.7559	0.9480
0.0422	8.2057	7500	0.3046	0.7594	0.7340	0.7465	0.9459
0.0422	8.3151	7600	0.3113	0.7640	0.7332	0.7483	0.9458
0.0422	8.4245	7700	0.3002	0.7579	0.7394	0.7485	0.9461
0.0422	8.5339	7800	0.3173	0.7742	0.7321	0.7526	0.9464
0.0422	8.6433	7900	0.3084	0.7766	0.7334	0.7544	0.9467
0.041	8.7527	8000	0.3118	0.7829	0.7325	0.7569	0.9477
0.041	8.8621	8100	0.3145	0.7788	0.7389	0.7583	0.9473
0.041	8.9716	8200	0.3123	0.7788	0.7366	0.7571	0.9480
0.041	9.0810	8300	0.3088	0.7754	0.7398	0.7572	0.9476
0.041	9.1904	8400	0.3101	0.7804	0.7415	0.7604	0.9491
0.0323	9.2998	8500	0.3152	0.7829	0.7357	0.7785	0.9482
0.0323	9.4092	8600	0.3061	0.7734	0.7398	0.7762	0.9476
0.0323	9.5186	8700	0.3086	0.7636	0.7437	0.7735	0.9476
0.0323	9.6280	8800	0.3162	0.7723	0.7390	0.7553	0.9476
0.0323	9.7374	8900	0.3070	0.7605	0.7419	0.7811	0.9467
0.0357	9.8468	9000	0.3117	0.7824	0.7396	0.7804	0.9481
0.0357	9.9562	9100	0.3130	0.7750	0.7396	0.7869	0.9472
0.0357	10.0656	9200	0.3095	0.7673	0.7405	0.7837	0.9476
0.0357	10.1751	9300	0.3179	0.7868	0.7357	0.7804	0.9477
0.0357	10.2845	9400	0.3077	0.7645	0.7405	0.7823	0.9472
0.0359	10.3939	9500	0.3128	0.7798	0.7366	0.7876	0.9476
0.0359	10.5033	9600	0.3151	0.7784	0.7377	0.7875	0.9475
0.0359	10.6127	9700	0.3138	0.7744	0.7420	0.7879	0.9478
0.0359	10.7221	9800	0.3115	0.7688	0.7415	0.7849	0.9475
0.0359	10.8315	9900	0.3097	0.7673	0.7411	0.7840	0.9472
0.0301	10.9409	10000	0.3095	0.7674	0.7409	0.7989	0.9474

Framework versions

• Transformers 4.41.0.dev0
• Pytorch 2.2.1+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1