shaikhadil26/layoutlm-sroie

To Use The Model

from transformers import LayoutLMForTokenClassification, LayoutLMv2Processor

   from transformers import LayoutLMForTokenClassification, LayoutLMv2Processor

   model = LayoutLMForTokenClassification.from_pretrained("shaikhadil26/layoutlm-sroie")
   processor = LayoutLMv2Processor.from_pretrained("shaikhadil26/layoutlm-sroie")

layoutlm-sroie

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: 0.0732
• Address: {'precision': 0.9800577276305432, 'recall': 0.9813452443510247, 'f1': 0.9807010634107917, 'number': 3806}
• Company: {'precision': 0.9241157556270096, 'recall': 0.9862731640356898, 'f1': 0.9541832669322708, 'number': 1457}
• Date: {'precision': 0.9520383693045563, 'recall': 0.9706601466992665, 'f1': 0.9612590799031476, 'number': 409}
• Total: {'precision': 0.6638888888888889, 'recall': 0.6675977653631285, 'f1': 0.6657381615598885, 'number': 358}
• Overall Precision: 0.9455
• Overall Recall: 0.9632
• Overall F1: 0.9542
• Overall Accuracy: 0.9863

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	Address	Company	Date	Total	Overall Precision	Overall Recall	Overall F1	Overall Accuracy
0.4342	1.0	40	0.0876	{'precision': 0.974816369359916, 'recall': 0.976353126642144, 'f1': 0.9755841428196377, 'number': 3806}	{'precision': 0.8865598027127004, 'recall': 0.9869595058339052, 'f1': 0.9340695030854174, 'number': 1457}	{'precision': 0.8112449799196787, 'recall': 0.9877750611246944, 'f1': 0.8908489525909592, 'number': 409}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 358}	0.9370	0.9217	0.9293	0.9794
0.0586	2.0	80	0.0591	{'precision': 0.9761780104712042, 'recall': 0.9797687861271677, 'f1': 0.977970102281668, 'number': 3806}	{'precision': 0.9187301587301587, 'recall': 0.9931365820178449, 'f1': 0.954485488126649, 'number': 1457}	{'precision': 0.8975501113585747, 'recall': 0.9853300733496333, 'f1': 0.9393939393939394, 'number': 409}	{'precision': 0.7073170731707317, 'recall': 0.40502793296089384, 'f1': 0.5150976909413854, 'number': 358}	0.9463	0.9493	0.9478	0.9844
0.039	3.0	120	0.0569	{'precision': 0.975006508721687, 'recall': 0.9839726747241198, 'f1': 0.9794690728390218, 'number': 3806}	{'precision': 0.9282970550576184, 'recall': 0.9951956074124915, 'f1': 0.9605829744948658, 'number': 1457}	{'precision': 0.9232558139534883, 'recall': 0.9706601466992665, 'f1': 0.9463647199046483, 'number': 409}	{'precision': 0.6103542234332425, 'recall': 0.6256983240223464, 'f1': 0.6179310344827587, 'number': 358}	0.9381	0.9645	0.9511	0.9852
0.0312	4.0	160	0.0546	{'precision': 0.9765135699373695, 'recall': 0.9831844456121913, 'f1': 0.9798376538360828, 'number': 3806}	{'precision': 0.9276105060858424, 'recall': 0.9938229238160604, 'f1': 0.9595758780649436, 'number': 1457}	{'precision': 0.9473684210526315, 'recall': 0.9682151589242054, 'f1': 0.9576783555018137, 'number': 409}	{'precision': 0.6116504854368932, 'recall': 0.7039106145251397, 'f1': 0.6545454545454545, 'number': 358}	0.9381	0.9682	0.9529	0.9858
0.0246	5.0	200	0.0555	{'precision': 0.9772430028773215, 'recall': 0.9816079873883342, 'f1': 0.9794206317997116, 'number': 3806}	{'precision': 0.9298132646490663, 'recall': 0.9910775566231984, 'f1': 0.959468438538206, 'number': 1457}	{'precision': 0.9537712895377128, 'recall': 0.9584352078239609, 'f1': 0.9560975609756097, 'number': 409}	{'precision': 0.6434108527131783, 'recall': 0.6955307262569832, 'f1': 0.6684563758389261, 'number': 358}	0.9428	0.9653	0.9539	0.9861
0.0206	6.0	240	0.0531	{'precision': 0.9782893015956056, 'recall': 0.9826589595375722, 'f1': 0.9804692620264778, 'number': 3806}	{'precision': 0.9414088215931534, 'recall': 0.9814687714481812, 'f1': 0.9610215053763441, 'number': 1457}	{'precision': 0.9628712871287128, 'recall': 0.9511002444987775, 'f1': 0.956949569495695, 'number': 409}	{'precision': 0.6811594202898551, 'recall': 0.6564245810055865, 'f1': 0.6685633001422475, 'number': 358}	0.9512	0.9609	0.9560	0.9868
0.0166	7.0	280	0.0579	{'precision': 0.9767987486965589, 'recall': 0.9844981607987389, 'f1': 0.9806333420570532, 'number': 3806}	{'precision': 0.9272844272844273, 'recall': 0.9890185312285518, 'f1': 0.9571570906675522, 'number': 1457}	{'precision': 0.9562043795620438, 'recall': 0.960880195599022, 'f1': 0.9585365853658536, 'number': 409}	{'precision': 0.6685236768802229, 'recall': 0.6703910614525139, 'f1': 0.6694560669456067, 'number': 358}	0.9450	0.9653	0.9550	0.9865
0.014	8.0	320	0.0614	{'precision': 0.9785396493064643, 'recall': 0.9823962165002628, 'f1': 0.9804641405532976, 'number': 3806}	{'precision': 0.9277885235332044, 'recall': 0.9876458476321208, 'f1': 0.9567819148936171, 'number': 1457}	{'precision': 0.9585365853658536, 'recall': 0.960880195599022, 'f1': 0.9597069597069597, 'number': 409}	{'precision': 0.6553524804177546, 'recall': 0.7011173184357542, 'f1': 0.6774628879892037, 'number': 358}	0.9444	0.9655	0.9548	0.9865
0.0115	9.0	360	0.0642	{'precision': 0.9800524934383202, 'recall': 0.9810825013137152, 'f1': 0.9805672268907565, 'number': 3806}	{'precision': 0.9342875731945348, 'recall': 0.9855868222374743, 'f1': 0.9592518370073482, 'number': 1457}	{'precision': 0.9527186761229315, 'recall': 0.9853300733496333, 'f1': 0.9687500000000001, 'number': 409}	{'precision': 0.6483516483516484, 'recall': 0.659217877094972, 'f1': 0.6537396121883657, 'number': 358}	0.9470	0.9633	0.9551	0.9865
0.0103	10.0	400	0.0684	{'precision': 0.9808197582764057, 'recall': 0.9808197582764057, 'f1': 0.9808197582764057, 'number': 3806}	{'precision': 0.9224358974358975, 'recall': 0.9876458476321208, 'f1': 0.9539277427908518, 'number': 1457}	{'precision': 0.9519230769230769, 'recall': 0.9682151589242054, 'f1': 0.96, 'number': 409}	{'precision': 0.6473684210526316, 'recall': 0.6871508379888268, 'f1': 0.6666666666666667, 'number': 358}	0.9435	0.9642	0.9537	0.9861
0.0084	11.0	440	0.0704	{'precision': 0.981325618095739, 'recall': 0.9802942722017867, 'f1': 0.9808096740273397, 'number': 3806}	{'precision': 0.9265463917525774, 'recall': 0.9869595058339052, 'f1': 0.9557992688600864, 'number': 1457}	{'precision': 0.9519230769230769, 'recall': 0.9682151589242054, 'f1': 0.96, 'number': 409}	{'precision': 0.6497326203208557, 'recall': 0.6787709497206704, 'f1': 0.6639344262295083, 'number': 358}	0.9453	0.9632	0.9542	0.9863
0.0077	12.0	480	0.0704	{'precision': 0.9805672268907563, 'recall': 0.9810825013137152, 'f1': 0.9808247964276333, 'number': 3806}	{'precision': 0.931950745301361, 'recall': 0.9869595058339052, 'f1': 0.9586666666666667, 'number': 1457}	{'precision': 0.9544364508393285, 'recall': 0.9731051344743277, 'f1': 0.963680387409201, 'number': 409}	{'precision': 0.6764705882352942, 'recall': 0.6424581005586593, 'f1': 0.6590257879656161, 'number': 358}	0.9496	0.9619	0.9557	0.9867
0.0075	13.0	520	0.0728	{'precision': 0.9792976939203354, 'recall': 0.9818707304256438, 'f1': 0.9805825242718447, 'number': 3806}	{'precision': 0.9258542875564152, 'recall': 0.9855868222374743, 'f1': 0.9547872340425532, 'number': 1457}	{'precision': 0.9538834951456311, 'recall': 0.960880195599022, 'f1': 0.9573690621193666, 'number': 409}	{'precision': 0.6502732240437158, 'recall': 0.664804469273743, 'f1': 0.6574585635359116, 'number': 358}	0.9445	0.9625	0.9534	0.9861
0.0068	14.0	560	0.0732	{'precision': 0.9805723286951956, 'recall': 0.9813452443510247, 'f1': 0.9809586342744582, 'number': 3806}	{'precision': 0.9229287090558767, 'recall': 0.9862731640356898, 'f1': 0.953550099535501, 'number': 1457}	{'precision': 0.9520383693045563, 'recall': 0.9706601466992665, 'f1': 0.9612590799031476, 'number': 409}	{'precision': 0.667590027700831, 'recall': 0.6731843575418994, 'f1': 0.6703755215577191, 'number': 358}	0.9456	0.9635	0.9545	0.9864
0.0066	15.0	600	0.0732	{'precision': 0.9800577276305432, 'recall': 0.9813452443510247, 'f1': 0.9807010634107917, 'number': 3806}	{'precision': 0.9241157556270096, 'recall': 0.9862731640356898, 'f1': 0.9541832669322708, 'number': 1457}	{'precision': 0.9520383693045563, 'recall': 0.9706601466992665, 'f1': 0.9612590799031476, 'number': 409}	{'precision': 0.6638888888888889, 'recall': 0.6675977653631285, 'f1': 0.6657381615598885, 'number': 358}	0.9455	0.9632	0.9542	0.9863

Framework versions

• Transformers 4.46.2
• Pytorch 2.5.0+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3