LayoutLMv3_97_1

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

• Loss: 0.8446
• Precision: 0.5939
• Recall: 0.8376
• F1: 0.6950
• Accuracy: 0.8952

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: 2000

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	2.44	100	0.4463	0.4830	0.7265	0.5802	0.8599
No log	4.88	200	0.4064	0.5924	0.7949	0.6788	0.8884
No log	7.32	300	0.4774	0.5813	0.7949	0.6715	0.8907
No log	9.76	400	0.5800	0.6013	0.7863	0.6815	0.8907
0.2076	12.2	500	0.6426	0.6209	0.8120	0.7037	0.8952
0.2076	14.63	600	0.6872	0.5939	0.8376	0.6950	0.8907
0.2076	17.07	700	0.7801	0.5915	0.8291	0.6904	0.8918
0.2076	19.51	800	0.7865	0.5890	0.8205	0.6857	0.8895
0.2076	21.95	900	0.8533	0.5854	0.8205	0.6833	0.8895
0.0109	24.39	1000	0.7738	0.5864	0.8120	0.6810	0.8941
0.0109	26.83	1100	0.8297	0.5854	0.8205	0.6833	0.8872
0.0109	29.27	1200	0.7690	0.6062	0.8291	0.7004	0.8975
0.0109	31.71	1300	0.8629	0.5904	0.8376	0.6926	0.8895
0.0109	34.15	1400	0.8104	0.5976	0.8376	0.6975	0.8941
0.0027	36.59	1500	0.7864	0.5926	0.8205	0.6882	0.8929
0.0027	39.02	1600	0.8002	0.6037	0.8462	0.7046	0.8986
0.0027	41.46	1700	0.8049	0.5964	0.8462	0.6996	0.8964
0.0027	43.9	1800	0.8355	0.5939	0.8376	0.6950	0.8952
0.0027	46.34	1900	0.8402	0.5939	0.8376	0.6950	0.8952
0.001	48.78	2000	0.8446	0.5939	0.8376	0.6950	0.8952

Framework versions

• Transformers 4.29.2
• Pytorch 2.0.1+cu118
• Datasets 2.14.4
• Tokenizers 0.13.3