layoutlm-FUNSDxSynthetic-1fold

This model is a fine-tuned version of microsoft/layoutlm-base-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.7358
• Eader: {'precision': 0.4666666666666667, 'recall': 0.3373493975903614, 'f1': 0.3916083916083916, 'number': 83}
• Nswer: {'precision': 0.5065502183406113, 'recall': 0.5658536585365853, 'f1': 0.5345622119815668, 'number': 205}
• Uestion: {'precision': 0.43373493975903615, 'recall': 0.4675324675324675, 'f1': 0.45, 'number': 231}
• Overall Precision: 0.4684
• Overall Recall: 0.4855
• Overall F1: 0.4768
• Overall Accuracy: 0.7947

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: 3e-05
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Eader	Nswer	Uestion	Overall Precision	Overall Recall	Overall F1	Overall Accuracy
1.2347	1.0	12	1.0175	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 83}	{'precision': 0.12563667232597622, 'recall': 0.36097560975609755, 'f1': 0.18639798488664988, 'number': 205}	{'precision': 0.11375212224108659, 'recall': 0.29004329004329005, 'f1': 0.16341463414634144, 'number': 231}	0.1197	0.2717	0.1662	0.6488
0.9147	2.0	24	0.7456	{'precision': 0.07142857142857142, 'recall': 0.012048192771084338, 'f1': 0.02061855670103093, 'number': 83}	{'precision': 0.2911764705882353, 'recall': 0.48292682926829267, 'f1': 0.36330275229357795, 'number': 205}	{'precision': 0.24507042253521127, 'recall': 0.37662337662337664, 'f1': 0.29692832764505117, 'number': 231}	0.2638	0.3603	0.3046	0.7487
0.675	3.0	36	0.6233	{'precision': 0.19148936170212766, 'recall': 0.10843373493975904, 'f1': 0.13846153846153844, 'number': 83}	{'precision': 0.36551724137931035, 'recall': 0.5170731707317073, 'f1': 0.42828282828282827, 'number': 205}	{'precision': 0.32094594594594594, 'recall': 0.41125541125541126, 'f1': 0.36053130929791266, 'number': 231}	0.3318	0.4046	0.3646	0.7793
0.4928	4.0	48	0.6064	{'precision': 0.3584905660377358, 'recall': 0.2289156626506024, 'f1': 0.2794117647058823, 'number': 83}	{'precision': 0.42592592592592593, 'recall': 0.5609756097560976, 'f1': 0.4842105263157895, 'number': 205}	{'precision': 0.34172661870503596, 'recall': 0.41125541125541126, 'f1': 0.37328094302554027, 'number': 231}	0.3810	0.4412	0.4089	0.7733
0.4142	5.0	60	0.5817	{'precision': 0.4489795918367347, 'recall': 0.26506024096385544, 'f1': 0.33333333333333337, 'number': 83}	{'precision': 0.4573643410852713, 'recall': 0.5756097560975609, 'f1': 0.5097192224622029, 'number': 205}	{'precision': 0.3754646840148699, 'recall': 0.43722943722943725, 'f1': 0.404, 'number': 231}	0.4184	0.4644	0.4402	0.7878
0.3198	6.0	72	0.5888	{'precision': 0.39285714285714285, 'recall': 0.26506024096385544, 'f1': 0.3165467625899281, 'number': 83}	{'precision': 0.4674329501915709, 'recall': 0.5951219512195122, 'f1': 0.5236051502145922, 'number': 205}	{'precision': 0.381294964028777, 'recall': 0.4588744588744589, 'f1': 0.4165029469548133, 'number': 231}	0.4202	0.4817	0.4488	0.7998
0.2752	7.0	84	0.6608	{'precision': 0.3968253968253968, 'recall': 0.30120481927710846, 'f1': 0.34246575342465757, 'number': 83}	{'precision': 0.47639484978540775, 'recall': 0.5414634146341464, 'f1': 0.5068493150684932, 'number': 205}	{'precision': 0.40458015267175573, 'recall': 0.4588744588744589, 'f1': 0.4300202839756592, 'number': 231}	0.4337	0.4663	0.4494	0.7728
0.2275	8.0	96	0.6552	{'precision': 0.3924050632911392, 'recall': 0.37349397590361444, 'f1': 0.38271604938271603, 'number': 83}	{'precision': 0.47478991596638653, 'recall': 0.551219512195122, 'f1': 0.510158013544018, 'number': 205}	{'precision': 0.37593984962406013, 'recall': 0.4329004329004329, 'f1': 0.40241448692152915, 'number': 231}	0.4185	0.4701	0.4428	0.7848
0.1947	9.0	108	0.6603	{'precision': 0.4166666666666667, 'recall': 0.30120481927710846, 'f1': 0.34965034965034963, 'number': 83}	{'precision': 0.4769874476987448, 'recall': 0.5560975609756098, 'f1': 0.5135135135135136, 'number': 205}	{'precision': 0.3917910447761194, 'recall': 0.45454545454545453, 'f1': 0.42084168336673344, 'number': 231}	0.4303	0.4701	0.4494	0.7867
0.169	10.0	120	0.6796	{'precision': 0.4264705882352941, 'recall': 0.3493975903614458, 'f1': 0.3841059602649007, 'number': 83}	{'precision': 0.4915254237288136, 'recall': 0.5658536585365853, 'f1': 0.5260770975056689, 'number': 205}	{'precision': 0.39015151515151514, 'recall': 0.4458874458874459, 'f1': 0.4161616161616162, 'number': 231}	0.4366	0.4778	0.4563	0.7891
0.1462	11.0	132	0.6880	{'precision': 0.41379310344827586, 'recall': 0.2891566265060241, 'f1': 0.3404255319148936, 'number': 83}	{'precision': 0.5109170305676856, 'recall': 0.5707317073170731, 'f1': 0.5391705069124425, 'number': 205}	{'precision': 0.41832669322709165, 'recall': 0.45454545454545453, 'f1': 0.43568464730290457, 'number': 231}	0.4572	0.4740	0.4655	0.7949
0.1295	12.0	144	0.7007	{'precision': 0.43283582089552236, 'recall': 0.3493975903614458, 'f1': 0.38666666666666666, 'number': 83}	{'precision': 0.4936708860759494, 'recall': 0.5707317073170731, 'f1': 0.5294117647058822, 'number': 205}	{'precision': 0.42023346303501946, 'recall': 0.4675324675324675, 'f1': 0.4426229508196721, 'number': 231}	0.4528	0.4894	0.4704	0.7951
0.1208	13.0	156	0.7216	{'precision': 0.4727272727272727, 'recall': 0.3132530120481928, 'f1': 0.3768115942028985, 'number': 83}	{'precision': 0.4957983193277311, 'recall': 0.5756097560975609, 'f1': 0.5327313769751693, 'number': 205}	{'precision': 0.42248062015503873, 'recall': 0.47186147186147187, 'f1': 0.44580777096114516, 'number': 231}	0.4592	0.4875	0.4729	0.7974
0.1146	14.0	168	0.7440	{'precision': 0.45901639344262296, 'recall': 0.3373493975903614, 'f1': 0.38888888888888884, 'number': 83}	{'precision': 0.5066666666666667, 'recall': 0.5560975609756098, 'f1': 0.5302325581395348, 'number': 205}	{'precision': 0.4291497975708502, 'recall': 0.4588744588744589, 'f1': 0.4435146443514644, 'number': 231}	0.4653	0.4778	0.4715	0.7910
0.1083	15.0	180	0.7358	{'precision': 0.4666666666666667, 'recall': 0.3373493975903614, 'f1': 0.3916083916083916, 'number': 83}	{'precision': 0.5065502183406113, 'recall': 0.5658536585365853, 'f1': 0.5345622119815668, 'number': 205}	{'precision': 0.43373493975903615, 'recall': 0.4675324675324675, 'f1': 0.45, 'number': 231}	0.4684	0.4855	0.4768	0.7947

Framework versions

• Transformers 4.49.0
• Pytorch 2.6.0+cu124
• Datasets 3.3.2
• Tokenizers 0.21.0