metadata
license: mit
base_model: microsoft/layoutlm-base-uncased
tags:
- generated_from_trainer
model-index:
- name: layoutlm-custom_no_text
results: []
layoutlm-custom_no_text
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: 0.3594
- Noise: {'precision': 0.5976627712854758, 'recall': 0.5700636942675159, 'f1': 0.5835370823145885, 'number': 628}
- Signal: {'precision': 0.5559265442404007, 'recall': 0.5302547770700637, 'f1': 0.5427872860635697, 'number': 628}
- Overall Precision: 0.5768
- Overall Recall: 0.5502
- Overall F1: 0.5632
- Overall Accuracy: 0.8777
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: 8
- eval_batch_size: 8
- 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 | Noise | Signal | Overall Precision | Overall Recall | Overall F1 | Overall Accuracy |
|---|---|---|---|---|---|---|---|---|---|
| 0.5172 | 1.0 | 18 | 0.4915 | {'precision': 0.3973288814691152, 'recall': 0.37898089171974525, 'f1': 0.3879380603096985, 'number': 628} | {'precision': 0.36894824707846413, 'recall': 0.3519108280254777, 'f1': 0.3602281988590057, 'number': 628} | 0.3831 | 0.3654 | 0.3741 | 0.7779 |
| 0.4057 | 2.0 | 36 | 0.4306 | {'precision': 0.42788461538461536, 'recall': 0.4251592356687898, 'f1': 0.426517571884984, 'number': 628} | {'precision': 0.3766025641025641, 'recall': 0.37420382165605093, 'f1': 0.3753993610223642, 'number': 628} | 0.4022 | 0.3997 | 0.4010 | 0.8151 |
| 0.3616 | 3.0 | 54 | 0.4145 | {'precision': 0.444633730834753, 'recall': 0.4156050955414013, 'f1': 0.4296296296296297, 'number': 628} | {'precision': 0.41056218057921634, 'recall': 0.3837579617834395, 'f1': 0.3967078189300412, 'number': 628} | 0.4276 | 0.3997 | 0.4132 | 0.8237 |
| 0.3278 | 4.0 | 72 | 0.3994 | {'precision': 0.5050167224080268, 'recall': 0.48089171974522293, 'f1': 0.4926590538336052, 'number': 628} | {'precision': 0.4698996655518395, 'recall': 0.44745222929936307, 'f1': 0.4584013050570963, 'number': 628} | 0.4875 | 0.4642 | 0.4755 | 0.8366 |
| 0.2966 | 5.0 | 90 | 0.3795 | {'precision': 0.5129533678756477, 'recall': 0.4729299363057325, 'f1': 0.49212924606462305, 'number': 628} | {'precision': 0.5043177892918825, 'recall': 0.46496815286624205, 'f1': 0.48384424192212094, 'number': 628} | 0.5086 | 0.4689 | 0.4880 | 0.8489 |
| 0.2717 | 6.0 | 108 | 0.3526 | {'precision': 0.5459272097053726, 'recall': 0.5015923566878981, 'f1': 0.5228215767634854, 'number': 628} | {'precision': 0.511265164644714, 'recall': 0.4697452229299363, 'f1': 0.48962655601659755, 'number': 628} | 0.5286 | 0.4857 | 0.5062 | 0.8581 |
| 0.2441 | 7.0 | 126 | 0.3400 | {'precision': 0.5338208409506399, 'recall': 0.46496815286624205, 'f1': 0.4970212765957447, 'number': 628} | {'precision': 0.49725776965265084, 'recall': 0.43312101910828027, 'f1': 0.4629787234042553, 'number': 628} | 0.5155 | 0.4490 | 0.4800 | 0.8598 |
| 0.224 | 8.0 | 144 | 0.3324 | {'precision': 0.563922942206655, 'recall': 0.5127388535031847, 'f1': 0.5371142618849041, 'number': 628} | {'precision': 0.5288966725043783, 'recall': 0.48089171974522293, 'f1': 0.5037531276063387, 'number': 628} | 0.5464 | 0.4968 | 0.5204 | 0.8673 |
| 0.2044 | 9.0 | 162 | 0.3249 | {'precision': 0.5833333333333334, 'recall': 0.535031847133758, 'f1': 0.558139534883721, 'number': 628} | {'precision': 0.5347222222222222, 'recall': 0.49044585987261147, 'f1': 0.5116279069767441, 'number': 628} | 0.5590 | 0.5127 | 0.5349 | 0.8726 |
| 0.1914 | 10.0 | 180 | 0.3481 | {'precision': 0.5597920277296361, 'recall': 0.5143312101910829, 'f1': 0.5360995850622408, 'number': 628} | {'precision': 0.511265164644714, 'recall': 0.4697452229299363, 'f1': 0.48962655601659755, 'number': 628} | 0.5355 | 0.4920 | 0.5129 | 0.8645 |
| 0.1823 | 11.0 | 198 | 0.3412 | {'precision': 0.5963756177924218, 'recall': 0.5764331210191083, 'f1': 0.5862348178137652, 'number': 628} | {'precision': 0.5667215815485996, 'recall': 0.5477707006369427, 'f1': 0.557085020242915, 'number': 628} | 0.5815 | 0.5621 | 0.5717 | 0.8810 |
| 0.1672 | 12.0 | 216 | 0.3496 | {'precision': 0.5791245791245792, 'recall': 0.5477707006369427, 'f1': 0.563011456628478, 'number': 628} | {'precision': 0.5420875420875421, 'recall': 0.5127388535031847, 'f1': 0.5270049099836335, 'number': 628} | 0.5606 | 0.5303 | 0.5450 | 0.8735 |
| 0.1627 | 13.0 | 234 | 0.3675 | {'precision': 0.5953565505804311, 'recall': 0.571656050955414, 'f1': 0.5832656376929325, 'number': 628} | {'precision': 0.5621890547263682, 'recall': 0.5398089171974523, 'f1': 0.5507717303005686, 'number': 628} | 0.5788 | 0.5557 | 0.5670 | 0.8779 |
| 0.1592 | 14.0 | 252 | 0.3562 | {'precision': 0.596694214876033, 'recall': 0.5748407643312102, 'f1': 0.5855636658556367, 'number': 628} | {'precision': 0.5504132231404959, 'recall': 0.5302547770700637, 'f1': 0.5401459854014599, 'number': 628} | 0.5736 | 0.5525 | 0.5629 | 0.8757 |
| 0.1553 | 15.0 | 270 | 0.3594 | {'precision': 0.5976627712854758, 'recall': 0.5700636942675159, 'f1': 0.5835370823145885, 'number': 628} | {'precision': 0.5559265442404007, 'recall': 0.5302547770700637, 'f1': 0.5427872860635697, 'number': 628} | 0.5768 | 0.5502 | 0.5632 | 0.8777 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.0