ihebMissaoui/layoutlmv3-fine-tuned-funsd-kie

test

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

• Loss: 0.8476
• Precision: 0.8955
• Recall: 0.9071
• F1: 0.9013
• Accuracy: 0.8691

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
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	75	0.9234	0.6139	0.7526	0.6762	0.7416
No log	2.0	150	0.6101	0.7549	0.8341	0.7925	0.7940
No log	3.0	225	0.5135	0.8332	0.8882	0.8598	0.8091
No log	4.0	300	0.5467	0.8189	0.8624	0.8401	0.8109
No log	5.0	375	0.4879	0.8660	0.9051	0.8851	0.8504
No log	6.0	450	0.5352	0.8787	0.9180	0.8980	0.8480
0.5752	7.0	525	0.5900	0.8730	0.8982	0.8854	0.8343
0.5752	8.0	600	0.6014	0.8832	0.9016	0.8923	0.8506
0.5752	9.0	675	0.6173	0.8883	0.9126	0.9003	0.8538
0.5752	10.0	750	0.6278	0.8787	0.9141	0.8960	0.8571
0.5752	11.0	825	0.6573	0.8612	0.9155	0.8876	0.8326
0.5752	12.0	900	0.7333	0.8818	0.9006	0.8911	0.8387
0.5752	13.0	975	0.7489	0.8888	0.9136	0.9010	0.8502
0.1263	14.0	1050	0.7719	0.8908	0.8997	0.8952	0.8318
0.1263	15.0	1125	0.8295	0.8945	0.9101	0.9022	0.8438
0.1263	16.0	1200	0.8447	0.8798	0.9126	0.8959	0.8465
0.1263	17.0	1275	0.8359	0.9090	0.8932	0.9010	0.8486
0.1263	18.0	1350	0.8430	0.8966	0.9091	0.9028	0.8414
0.1263	19.0	1425	0.8179	0.8854	0.9021	0.8937	0.8400
0.0482	20.0	1500	0.8950	0.8968	0.8982	0.8975	0.8475
0.0482	21.0	1575	0.8790	0.9053	0.9121	0.9087	0.8565
0.0482	22.0	1650	0.7915	0.9056	0.9101	0.9078	0.8595
0.0482	23.0	1725	0.8760	0.8938	0.8952	0.8945	0.8504
0.0482	24.0	1800	0.8320	0.9113	0.9086	0.9100	0.8625
0.0482	25.0	1875	0.8880	0.9017	0.9021	0.9019	0.8538
0.0482	26.0	1950	0.8611	0.9083	0.9101	0.9092	0.8499
0.0163	27.0	2025	0.8747	0.9068	0.9086	0.9077	0.8600
0.0163	28.0	2100	0.8476	0.8955	0.9071	0.9013	0.8691

Framework versions

• Transformers 4.42.0.dev0
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1