lilt-en-funsd
This model is a fine-tuned version of SCUT-DLVCLab/lilt-roberta-en-base on the funsd-layoutlmv3 dataset. It achieves the following results on the evaluation set:
- Loss: 1.5496
- Answer: {'precision': 0.875, 'recall': 0.9253365973072215, 'f1': 0.8994646044021416, 'number': 817}
- Header: {'precision': 0.6276595744680851, 'recall': 0.4957983193277311, 'f1': 0.5539906103286385, 'number': 119}
- Question: {'precision': 0.9049360146252285, 'recall': 0.9192200557103064, 'f1': 0.9120221096269001, 'number': 1077}
- Overall Precision: 0.8796
- Overall Recall: 0.8967
- Overall F1: 0.8881
- Overall Accuracy: 0.8134
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: 5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- training_steps: 2500
Training results
| Training Loss | Epoch | Step | Validation Loss | Answer | Header | Question | Overall Precision | Overall Recall | Overall F1 | Overall Accuracy |
|---|---|---|---|---|---|---|---|---|---|---|
| 0.434 | 10.53 | 200 | 1.0227 | {'precision': 0.8357705286839145, 'recall': 0.9094247246022031, 'f1': 0.8710433763188746, 'number': 817} | {'precision': 0.7058823529411765, 'recall': 0.40336134453781514, 'f1': 0.5133689839572192, 'number': 119} | {'precision': 0.8683522231909329, 'recall': 0.924791086350975, 'f1': 0.89568345323741, 'number': 1077} | 0.8493 | 0.8877 | 0.8681 | 0.7935 |
| 0.0484 | 21.05 | 400 | 1.3626 | {'precision': 0.8098360655737705, 'recall': 0.9069767441860465, 'f1': 0.8556581986143187, 'number': 817} | {'precision': 0.6086956521739131, 'recall': 0.47058823529411764, 'f1': 0.5308056872037914, 'number': 119} | {'precision': 0.8613333333333333, 'recall': 0.8997214484679665, 'f1': 0.8801089918256131, 'number': 1077} | 0.8283 | 0.8773 | 0.8521 | 0.7995 |
| 0.0168 | 31.58 | 600 | 1.3003 | {'precision': 0.8440046565774156, 'recall': 0.8873929008567931, 'f1': 0.8651551312649164, 'number': 817} | {'precision': 0.6421052631578947, 'recall': 0.5126050420168067, 'f1': 0.5700934579439252, 'number': 119} | {'precision': 0.8776595744680851, 'recall': 0.9192200557103064, 'f1': 0.8979591836734694, 'number': 1077} | 0.8530 | 0.8823 | 0.8674 | 0.8189 |
| 0.008 | 42.11 | 800 | 1.3225 | {'precision': 0.8584795321637427, 'recall': 0.8984088127294981, 'f1': 0.8779904306220095, 'number': 817} | {'precision': 0.5736434108527132, 'recall': 0.6218487394957983, 'f1': 0.596774193548387, 'number': 119} | {'precision': 0.888468809073724, 'recall': 0.872794800371402, 'f1': 0.8805620608899298, 'number': 1077} | 0.8560 | 0.8684 | 0.8621 | 0.8210 |
| 0.0059 | 52.63 | 1000 | 1.6362 | {'precision': 0.8307522123893806, 'recall': 0.9192166462668299, 'f1': 0.8727484020918072, 'number': 817} | {'precision': 0.6419753086419753, 'recall': 0.4369747899159664, 'f1': 0.52, 'number': 119} | {'precision': 0.8944444444444445, 'recall': 0.8969359331476323, 'f1': 0.8956884561891516, 'number': 1077} | 0.8567 | 0.8788 | 0.8676 | 0.8061 |
| 0.0027 | 63.16 | 1200 | 1.6927 | {'precision': 0.8269858541893362, 'recall': 0.9302325581395349, 'f1': 0.8755760368663594, 'number': 817} | {'precision': 0.6046511627906976, 'recall': 0.4369747899159664, 'f1': 0.5073170731707317, 'number': 119} | {'precision': 0.9000925069380203, 'recall': 0.903435468895079, 'f1': 0.901760889712697, 'number': 1077} | 0.8557 | 0.8867 | 0.8709 | 0.7939 |
| 0.002 | 73.68 | 1400 | 1.4609 | {'precision': 0.8479467258601554, 'recall': 0.9351285189718482, 'f1': 0.889406286379511, 'number': 817} | {'precision': 0.5726495726495726, 'recall': 0.5630252100840336, 'f1': 0.5677966101694915, 'number': 119} | {'precision': 0.8917431192660551, 'recall': 0.9025069637883009, 'f1': 0.8970927549607752, 'number': 1077} | 0.8553 | 0.8957 | 0.8750 | 0.7965 |
| 0.0012 | 84.21 | 1600 | 1.4851 | {'precision': 0.865909090909091, 'recall': 0.9326805385556916, 'f1': 0.8980553918680023, 'number': 817} | {'precision': 0.6074766355140186, 'recall': 0.5462184873949579, 'f1': 0.575221238938053, 'number': 119} | {'precision': 0.9008341056533827, 'recall': 0.9025069637883009, 'f1': 0.901669758812616, 'number': 1077} | 0.8708 | 0.8937 | 0.8821 | 0.8131 |
| 0.0006 | 94.74 | 1800 | 1.5228 | {'precision': 0.850613154960981, 'recall': 0.9339045287637698, 'f1': 0.8903150525087514, 'number': 817} | {'precision': 0.594059405940594, 'recall': 0.5042016806722689, 'f1': 0.5454545454545453, 'number': 119} | {'precision': 0.896709323583181, 'recall': 0.9108635097493036, 'f1': 0.9037309995393827, 'number': 1077} | 0.8623 | 0.8962 | 0.8789 | 0.8082 |
| 0.0004 | 105.26 | 2000 | 1.5287 | {'precision': 0.867579908675799, 'recall': 0.9302325581395349, 'f1': 0.8978145304193739, 'number': 817} | {'precision': 0.6222222222222222, 'recall': 0.47058823529411764, 'f1': 0.5358851674641149, 'number': 119} | {'precision': 0.8917710196779964, 'recall': 0.9257195914577531, 'f1': 0.9084282460136676, 'number': 1077} | 0.8700 | 0.9006 | 0.8850 | 0.8128 |
| 0.0003 | 115.79 | 2200 | 1.5306 | {'precision': 0.8766006984866124, 'recall': 0.9216646266829865, 'f1': 0.8985680190930787, 'number': 817} | {'precision': 0.6263736263736264, 'recall': 0.4789915966386555, 'f1': 0.5428571428571428, 'number': 119} | {'precision': 0.8902765388046388, 'recall': 0.9266480965645311, 'f1': 0.908098271155596, 'number': 1077} | 0.8730 | 0.8982 | 0.8854 | 0.8127 |
| 0.0001 | 126.32 | 2400 | 1.5496 | {'precision': 0.875, 'recall': 0.9253365973072215, 'f1': 0.8994646044021416, 'number': 817} | {'precision': 0.6276595744680851, 'recall': 0.4957983193277311, 'f1': 0.5539906103286385, 'number': 119} | {'precision': 0.9049360146252285, 'recall': 0.9192200557103064, 'f1': 0.9120221096269001, 'number': 1077} | 0.8796 | 0.8967 | 0.8881 | 0.8134 |
Framework versions
- Transformers 4.30.2
- Pytorch 2.0.1+cu118
- Datasets 2.13.0
- Tokenizers 0.13.3
- Downloads last month
- 6
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.