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: 3.6446

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: 10

Training results

Training Loss	Epoch	Step	Validation Loss
5.1899	0.2212	50	4.5198
4.3841	0.4425	100	3.9910
4.0167	0.6637	150	3.9030
3.7906	0.8850	200	3.4974
3.426	1.1062	250	3.7834
3.0774	1.3274	300	3.2013
2.9541	1.5487	350	3.0711
2.7072	1.7699	400	2.7067
2.3127	1.9912	450	2.6800
2.0786	2.2124	500	2.6677
1.9055	2.4336	550	2.5482
2.0272	2.6549	600	2.2344
1.6753	2.8761	650	2.3265
1.4848	3.0973	700	2.5170
1.4359	3.3186	750	2.4527
1.2884	3.5398	800	2.4033
1.3217	3.7611	850	2.0981
1.3359	3.9823	900	2.2481
0.9068	4.2035	950	2.4053
1.1537	4.4248	1000	2.5739
0.8742	4.6460	1050	2.5003
0.9135	4.8673	1100	2.5511
0.9073	5.0885	1150	2.6724
0.6596	5.3097	1200	2.6174
0.7797	5.5310	1250	2.8350
0.6031	5.7522	1300	3.1642
0.7287	5.9735	1350	3.0317
0.4586	6.1947	1400	3.1821
0.5755	6.4159	1450	3.1222
0.4174	6.6372	1500	3.6997
0.2839	6.8584	1550	3.7044
0.5378	7.0796	1600	3.4223
0.3549	7.3009	1650	3.3740
0.2239	7.5221	1700	3.6018
0.3209	7.7434	1750	3.3689
0.2594	7.9646	1800	3.6625
0.1606	8.1858	1850	3.6916
0.1525	8.4071	1900	3.6299
0.1104	8.6283	1950	3.7133
0.3046	8.8496	2000	3.7701
0.3161	9.0708	2050	3.6224
0.1331	9.2920	2100	3.6198
0.2595	9.5133	2150	3.6251
0.1928	9.7345	2200	3.6359
0.1465	9.9558	2250	3.6446

Framework versions

• Transformers 4.42.4
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1