layoutlm-funsd

alexneakameni/layoutlm-funsd

layoutlm-funsd is a token classification model from alexneakameni. Use it when you need labels on individual words, such as names. It is set up for transformers.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

19% of all-time downloads

All-time downloads

Public

Parameters

113M

1.4 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.bin451 MB · 50%

Parameter types

How the weights are stored.

F32113M · 100%

Try a prompt

Base models

microsoft/layoutlm-base-uncased

Datasets

funsd

Model card

From the Hugging Face model README

layoutlm-funsd

This model is a fine-tuned version of microsoft/layoutlm-base-uncased on the funsd dataset. It achieves the following results on the evaluation set:

Loss: 0.7960
Answer: {'precision': 0.7169603524229075, 'recall': 0.8046971569839307, 'f1': 0.7582993593476993, 'number': 809}
Header: {'precision': 0.36619718309859156, 'recall': 0.4369747899159664, 'f1': 0.39846743295019166, 'number': 119}
Question: {'precision': 0.7883408071748879, 'recall': 0.8253521126760563, 'f1': 0.8064220183486238, 'number': 1065}
Overall Precision: 0.7307
Overall Recall: 0.7938
Overall F1: 0.7609
Overall Accuracy: 0.8081

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: 6
eval_batch_size: 4
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 15
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Answer	Header	Question	Overall Precision	Overall Recall	Overall F1	Overall Accuracy
1.6386	1.0	25	1.2949	{'precision': 0.08352668213457076, 'recall': 0.08899876390605686, 'f1': 0.08617594254937162, 'number': 809}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 119}	{'precision': 0.36874571624400276, 'recall': 0.5051643192488263, 'f1': 0.42630744849445323, 'number': 1065}	0.2628	0.3061	0.2828	0.5116
1.0433	2.0	50	0.8005	{'precision': 0.5965447154471545, 'recall': 0.7255871446229913, 'f1': 0.6547685443390964, 'number': 809}	{'precision': 0.1111111111111111, 'recall': 0.058823529411764705, 'f1': 0.07692307692307691, 'number': 119}	{'precision': 0.6574487065120428, 'recall': 0.692018779342723, 'f1': 0.6742909423604757, 'number': 1065}	0.6139	0.6678	0.6398	0.7293
0.6891	3.0	75	0.6695	{'precision': 0.6335650446871897, 'recall': 0.788627935723115, 'f1': 0.7026431718061674, 'number': 809}	{'precision': 0.3246753246753247, 'recall': 0.21008403361344538, 'f1': 0.25510204081632654, 'number': 119}	{'precision': 0.7085862966175195, 'recall': 0.7671361502347418, 'f1': 0.7366997294860236, 'number': 1065}	0.6616	0.7426	0.6998	0.7752
0.532	4.0	100	0.6270	{'precision': 0.6573787409700722, 'recall': 0.7873918417799752, 'f1': 0.7165354330708661, 'number': 809}	{'precision': 0.2361111111111111, 'recall': 0.2857142857142857, 'f1': 0.25855513307984795, 'number': 119}	{'precision': 0.7153284671532847, 'recall': 0.828169014084507, 'f1': 0.7676240208877285, 'number': 1065}	0.6620	0.7792	0.7158	0.7961
0.4184	5.0	125	0.6174	{'precision': 0.6837160751565762, 'recall': 0.8096415327564895, 'f1': 0.7413695529145445, 'number': 809}	{'precision': 0.3063063063063063, 'recall': 0.2857142857142857, 'f1': 0.2956521739130435, 'number': 119}	{'precision': 0.7734657039711191, 'recall': 0.8046948356807512, 'f1': 0.7887712839392544, 'number': 1065}	0.7102	0.7757	0.7415	0.8025
0.3264	6.0	150	0.6493	{'precision': 0.6905537459283387, 'recall': 0.7861557478368356, 'f1': 0.7352601156069365, 'number': 809}	{'precision': 0.310126582278481, 'recall': 0.4117647058823529, 'f1': 0.35379061371841153, 'number': 119}	{'precision': 0.7713523131672598, 'recall': 0.8140845070422535, 'f1': 0.7921425308359983, 'number': 1065}	0.7045	0.7787	0.7398	0.8008
0.2661	7.0	175	0.6587	{'precision': 0.6857440166493236, 'recall': 0.8145859085290482, 'f1': 0.7446327683615819, 'number': 809}	{'precision': 0.32575757575757575, 'recall': 0.36134453781512604, 'f1': 0.3426294820717131, 'number': 119}	{'precision': 0.7720970537261699, 'recall': 0.8366197183098592, 'f1': 0.8030644434429923, 'number': 1065}	0.7089	0.7993	0.7514	0.8038
0.2246	8.0	200	0.7115	{'precision': 0.7111356119073869, 'recall': 0.7972805933250927, 'f1': 0.7517482517482517, 'number': 809}	{'precision': 0.2983425414364641, 'recall': 0.453781512605042, 'f1': 0.36, 'number': 119}	{'precision': 0.7891402714932126, 'recall': 0.8187793427230047, 'f1': 0.8036866359447005, 'number': 1065}	0.7164	0.7883	0.7506	0.8074
0.1928	9.0	225	0.7130	{'precision': 0.7094668117519043, 'recall': 0.8059332509270705, 'f1': 0.7546296296296295, 'number': 809}	{'precision': 0.3178294573643411, 'recall': 0.3445378151260504, 'f1': 0.33064516129032256, 'number': 119}	{'precision': 0.7908025247971145, 'recall': 0.8234741784037559, 'f1': 0.8068077276908925, 'number': 1065}	0.7279	0.7878	0.7566	0.8042
0.1598	10.0	250	0.7375	{'precision': 0.7242937853107345, 'recall': 0.792336217552534, 'f1': 0.756788665879575, 'number': 809}	{'precision': 0.375, 'recall': 0.42857142857142855, 'f1': 0.39999999999999997, 'number': 119}	{'precision': 0.788858939802336, 'recall': 0.8244131455399061, 'f1': 0.8062442607897153, 'number': 1065}	0.7357	0.7878	0.7608	0.8099
0.1444	11.0	275	0.7719	{'precision': 0.7027896995708155, 'recall': 0.8096415327564895, 'f1': 0.7524411257897761, 'number': 809}	{'precision': 0.34814814814814815, 'recall': 0.3949579831932773, 'f1': 0.3700787401574803, 'number': 119}	{'precision': 0.7825311942959001, 'recall': 0.8244131455399061, 'f1': 0.8029263831732967, 'number': 1065}	0.7218	0.7928	0.7556	0.8008
0.1251	12.0	300	0.7758	{'precision': 0.7133479212253829, 'recall': 0.8059332509270705, 'f1': 0.7568195008705745, 'number': 809}	{'precision': 0.38095238095238093, 'recall': 0.40336134453781514, 'f1': 0.39183673469387753, 'number': 119}	{'precision': 0.7880434782608695, 'recall': 0.8169014084507042, 'f1': 0.8022130013831259, 'number': 1065}	0.7323	0.7878	0.7590	0.8077
0.1124	13.0	325	0.7878	{'precision': 0.7150776053215078, 'recall': 0.7972805933250927, 'f1': 0.7539450613676213, 'number': 809}	{'precision': 0.38848920863309355, 'recall': 0.453781512605042, 'f1': 0.4186046511627907, 'number': 119}	{'precision': 0.7922312556458898, 'recall': 0.8234741784037559, 'f1': 0.8075506445672191, 'number': 1065}	0.7337	0.7908	0.7612	0.8094
0.1077	14.0	350	0.7945	{'precision': 0.7136612021857923, 'recall': 0.8071693448702101, 'f1': 0.7575406032482598, 'number': 809}	{'precision': 0.36619718309859156, 'recall': 0.4369747899159664, 'f1': 0.39846743295019166, 'number': 119}	{'precision': 0.7887197851387645, 'recall': 0.8272300469483568, 'f1': 0.8075160403299725, 'number': 1065}	0.7295	0.7958	0.7612	0.8098
0.1001	15.0	375	0.7960	{'precision': 0.7169603524229075, 'recall': 0.8046971569839307, 'f1': 0.7582993593476993, 'number': 809}	{'precision': 0.36619718309859156, 'recall': 0.4369747899159664, 'f1': 0.39846743295019166, 'number': 119}	{'precision': 0.7883408071748879, 'recall': 0.8253521126760563, 'f1': 0.8064220183486238, 'number': 1065}	0.7307	0.7938	0.7609	0.8081

Framework versions

Transformers 4.28.1
Pytorch 2.0.0+cu117
Datasets 2.11.0
Tokenizers 0.13.3

Task: Token Classification
Library: transformers
Type: layoutlm
Created: Apr 18, 2023
Updated: Sep 12, 2023