|
--- |
|
tags: |
|
- generated_from_trainer |
|
model-index: |
|
- name: icdar23-entrydetector_jointlabelledtext_breaks_indents_left_diff_right_ref |
|
results: [] |
|
--- |
|
|
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You |
|
should probably proofread and complete it, then remove this comment. --> |
|
|
|
# icdar23-entrydetector_jointlabelledtext_breaks_indents_left_diff_right_ref |
|
|
|
This model is a fine-tuned version of [HueyNemud/das22-10-camembert_pretrained](https://huggingface.co/HueyNemud/das22-10-camembert_pretrained) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.2611 |
|
- Act: {'precision': 0.7855491329479769, 'recall': 0.8905635648754915, 'f1': 0.8347665847665848, 'number': 1526} |
|
- Cardinal: {'precision': 0.9609375, 'recall': 0.9624413145539906, 'f1': 0.9616888193901486, 'number': 2556} |
|
- Cardinal+i-eend: {'precision': 1.0, 'recall': 0.2631578947368421, 'f1': 0.4166666666666667, 'number': 114} |
|
- Ft: {'precision': 0.3125, 'recall': 0.23809523809523808, 'f1': 0.27027027027027023, 'number': 21} |
|
- Loc: {'precision': 0.9030707610146862, 'recall': 0.9397054737427063, 'f1': 0.9210239651416122, 'number': 3599} |
|
- Loc+i-eend: {'precision': 0.9444444444444444, 'recall': 0.3617021276595745, 'f1': 0.5230769230769231, 'number': 47} |
|
- Per: {'precision': 0.915758896151053, 'recall': 0.9231332357247438, 'f1': 0.919431279620853, 'number': 2732} |
|
- Per+i-ebegin: {'precision': 0.9938223938223938, 'recall': 0.9877206446661551, 'f1': 0.9907621247113164, 'number': 2606} |
|
- Titre: {'precision': 0.6972972972972973, 'recall': 0.86, 'f1': 0.7701492537313434, 'number': 150} |
|
- Overall Precision: 0.9156 |
|
- Overall Recall: 0.9346 |
|
- Overall F1: 0.9250 |
|
- Overall Accuracy: 0.9418 |
|
|
|
## 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: 0.0001 |
|
- train_batch_size: 2 |
|
- eval_batch_size: 2 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- training_steps: 15000 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy | |
|
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:| |
|
| No log | 0.07 | 300 | 0.2539 | 0.8509 | 0.9103 | 0.8796 | 0.9523 | |
|
| 0.5632 | 0.14 | 600 | 0.1632 | 0.9225 | 0.9305 | 0.9265 | 0.9647 | |
|
| 0.5632 | 0.21 | 900 | 0.1571 | 0.9300 | 0.9345 | 0.9323 | 0.9638 | |
|
| 0.204 | 0.29 | 1200 | 0.1415 | 0.9322 | 0.9399 | 0.9361 | 0.9669 | |
|
| 0.1626 | 0.36 | 1500 | 0.1331 | 0.9428 | 0.9477 | 0.9452 | 0.9679 | |
|
| 0.1626 | 0.43 | 1800 | 0.1272 | 0.9384 | 0.9537 | 0.9460 | 0.9679 | |
|
| 0.1305 | 0.5 | 2100 | 0.1334 | 0.9435 | 0.9513 | 0.9474 | 0.9696 | |
|
| 0.1305 | 0.57 | 2400 | 0.1199 | 0.9410 | 0.9496 | 0.9452 | 0.9705 | |
|
| 0.1288 | 0.64 | 2700 | 0.1412 | 0.9401 | 0.9530 | 0.9465 | 0.9685 | |
|
| 0.1345 | 0.72 | 3000 | 0.1177 | 0.9407 | 0.9534 | 0.9470 | 0.9711 | |
|
| 0.1345 | 0.79 | 3300 | 0.1191 | 0.9417 | 0.9599 | 0.9507 | 0.9718 | |
|
| 0.1123 | 0.86 | 3600 | 0.1110 | 0.9472 | 0.9609 | 0.9540 | 0.9746 | |
|
| 0.1123 | 0.93 | 3900 | 0.1229 | 0.9343 | 0.9462 | 0.9402 | 0.9712 | |
|
| 0.1047 | 1.0 | 4200 | 0.1032 | 0.9521 | 0.9622 | 0.9571 | 0.9770 | |
|
| 0.0713 | 1.07 | 4500 | 0.1093 | 0.9343 | 0.9642 | 0.9490 | 0.9746 | |
|
| 0.0713 | 1.14 | 4800 | 0.1045 | 0.9499 | 0.9609 | 0.9554 | 0.9758 | |
|
| 0.0674 | 1.22 | 5100 | 0.1287 | 0.9382 | 0.9704 | 0.9541 | 0.9730 | |
|
| 0.0674 | 1.29 | 5400 | 0.0983 | 0.9520 | 0.9547 | 0.9533 | 0.9743 | |
|
| 0.0682 | 1.36 | 5700 | 0.1153 | 0.9468 | 0.9611 | 0.9539 | 0.9752 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.26.1 |
|
- Pytorch 1.13.1+cu116 |
|
- Datasets 2.9.0 |
|
- Tokenizers 0.13.2 |
|
|
