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: 3.8273

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: 8
• 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
2.0156	0.44	50	2.4679
1.8368	0.88	100	2.2079
1.5567	1.33	150	2.3312
1.3487	1.77	200	2.8410
1.2254	2.21	250	2.6996
1.1201	2.65	300	2.2915
0.8816	3.1	350	2.3419
0.7885	3.54	400	2.6410
0.7532	3.98	450	2.7539
0.5822	4.42	500	2.7213
0.5801	4.87	550	2.7429
0.5043	5.31	600	2.8523
0.4545	5.75	650	2.8666
0.4029	6.19	700	3.4559
0.3568	6.64	750	3.1760
0.3962	7.08	800	3.0625
0.2381	7.52	850	3.3868
0.2492	7.96	900	3.7453
0.3813	8.41	950	3.2516
0.2477	8.85	1000	3.4677
0.1834	9.29	1050	3.2748
0.2067	9.73	1100	3.7590
0.2062	10.18	1150	3.5956
0.1337	10.62	1200	3.8232
0.1785	11.06	1250	3.5264
0.0906	11.5	1300	3.6157
0.1649	11.95	1350	3.4667
0.1306	12.39	1400	3.7029
0.0529	12.83	1450	3.6307
0.0628	13.27	1500	3.5905
0.1015	13.72	1550	3.4659
0.0693	14.16	1600	3.7713
0.1111	14.6	1650	3.7680
0.0414	15.04	1700	3.8956
0.0256	15.49	1750	3.9021
0.0737	15.93	1800	3.9392
0.0577	16.37	1850	3.8129
0.0744	16.81	1900	3.8356
0.0698	17.26	1950	3.8406
0.0173	17.7	2000	3.8611
0.0667	18.14	2050	3.7995
0.0482	18.58	2100	3.8132
0.0458	19.03	2150	3.8335
0.0415	19.47	2200	3.8475
0.0236	19.91	2250	3.8273

Framework versions

• Transformers 4.34.1
• Pytorch 2.1.0+cu118
• Datasets 2.14.6
• Tokenizers 0.14.1