layoutlmv2-base-uncased_finetuned_docvqa

This model is a fine-tuned version of microsoft/layoutlmv2-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 5.5645

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: 4
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss
5.3224	0.2212	50	4.5586
4.5246	0.4425	100	4.1173
4.1619	0.6637	150	3.8601
3.7534	0.8850	200	3.6319
3.6105	1.1062	250	3.7778
3.3319	1.3274	300	3.1775
3.0645	1.5487	350	2.8592
2.8209	1.7699	400	2.7744
2.7174	1.9912	450	2.7408
2.0437	2.2124	500	2.7848
2.0063	2.4336	550	2.9319
1.9314	2.6549	600	2.3084
1.7939	2.8761	650	2.4124
1.7613	3.0973	700	2.5776
1.3099	3.3186	750	2.2375
1.4457	3.5398	800	2.7229
1.4964	3.7611	850	2.5109
1.428	3.9823	900	2.4552
0.9892	4.2035	950	3.2111
1.0568	4.4248	1000	2.3875
0.8754	4.6460	1050	2.8059
0.8201	4.8673	1100	2.5949
1.0239	5.0885	1150	2.8688
0.7348	5.3097	1200	2.8210
0.7866	5.5310	1250	2.4231
0.5954	5.7522	1300	2.8619
0.7299	5.9735	1350	2.8536
0.5132	6.1947	1400	2.6224
0.7035	6.4159	1450	3.2108
0.5626	6.6372	1500	2.8695
0.431	6.8584	1550	3.3508
0.4354	7.0796	1600	3.4196
0.3896	7.3009	1650	3.1219
0.4899	7.5221	1700	3.0649
0.5703	7.7434	1750	3.0621
0.435	7.9646	1800	3.3686
0.3251	8.1858	1850	3.2093
0.2464	8.4071	1900	3.9491
0.4524	8.6283	1950	3.4324
0.5715	8.8496	2000	3.5811
0.3552	9.0708	2050	3.9434
0.1147	9.2920	2100	4.5776
0.2613	9.5133	2150	4.0439
0.5679	9.7345	2200	3.4187
0.3372	9.9558	2250	3.3868
0.3143	10.1770	2300	4.2051
0.1989	10.3982	2350	3.7925
0.1859	10.6195	2400	4.1932
0.3882	10.8407	2450	4.1672
0.1824	11.0619	2500	4.3516
0.106	11.2832	2550	4.5112
0.2096	11.5044	2600	4.3784
0.1035	11.7257	2650	4.3866
0.2113	11.9469	2700	4.1279
0.2263	12.1681	2750	4.2749
0.1014	12.3894	2800	4.5176
0.1555	12.6106	2850	3.9479
0.1732	12.8319	2900	4.2414
0.1484	13.0531	2950	4.0296
0.1051	13.2743	3000	4.5086
0.1282	13.4956	3050	4.6194
0.1471	13.7168	3100	4.6707
0.1888	13.9381	3150	4.3906
0.0723	14.1593	3200	4.9790
0.0302	14.3805	3250	5.0363
0.1599	14.6018	3300	4.8371
0.1179	14.8230	3350	4.3327
0.1128	15.0442	3400	5.0618
0.0493	15.2655	3450	5.2469
0.0341	15.4867	3500	5.3640
0.0545	15.7080	3550	5.0736
0.0883	15.9292	3600	5.1372
0.0461	16.1504	3650	5.0354
0.0244	16.3717	3700	5.4353
0.0541	16.5929	3750	5.3114
0.0164	16.8142	3800	5.4107
0.0336	17.0354	3850	5.4258
0.0483	17.2566	3900	5.3555
0.0994	17.4779	3950	5.2090
0.0351	17.6991	4000	5.3768
0.0065	17.9204	4050	5.5076
0.0053	18.1416	4100	5.4823
0.0043	18.3628	4150	5.4850
0.0452	18.5841	4200	5.4849
0.0086	18.8053	4250	5.5881
0.0322	19.0265	4300	5.5167
0.0135	19.2478	4350	5.5502
0.0229	19.4690	4400	5.5385
0.042	19.6903	4450	5.5602
0.0404	19.9115	4500	5.5645

Framework versions

• Transformers 4.44.2
• Pytorch 2.5.0+cu121
• Datasets 3.1.0
• Tokenizers 0.19.1