icdar23-entrydetector_plaintext_breaks_indents_left_ref_right_ref
This model is a fine-tuned version of HueyNemud/das22-10-camembert_pretrained on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0068
- Ebegin: {'precision': 1.0, 'recall': 0.9793155321549455, 'f1': 0.9895496864905947, 'number': 2659}
- Eend: {'precision': 0.9988562714449104, 'recall': 0.9790732436472347, 'f1': 0.9888658237403285, 'number': 2676}
- Overall Precision: 0.9994
- Overall Recall: 0.9792
- Overall F1: 0.9892
- Overall Accuracy: 0.9983
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: 7500
Training results
Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
---|---|---|---|---|---|---|---|
No log | 0.07 | 300 | 0.0309 | 0.9637 | 0.9910 | 0.9771 | 0.9964 |
0.181 | 0.14 | 600 | 0.0144 | 0.9777 | 0.9863 | 0.9819 | 0.9974 |
0.181 | 0.21 | 900 | 0.0095 | 0.9969 | 0.9845 | 0.9906 | 0.9985 |
0.0168 | 0.29 | 1200 | 0.0105 | 0.9869 | 0.9913 | 0.9891 | 0.9982 |
0.011 | 0.36 | 1500 | 0.0063 | 0.9937 | 0.9915 | 0.9926 | 0.9988 |
0.011 | 0.43 | 1800 | 0.0064 | 0.9883 | 0.9940 | 0.9911 | 0.9986 |
0.01 | 0.5 | 2100 | 0.0203 | 0.9552 | 0.9507 | 0.9529 | 0.9922 |
0.01 | 0.57 | 2400 | 0.0049 | 0.9946 | 0.9925 | 0.9935 | 0.9989 |
0.0144 | 0.64 | 2700 | 0.0056 | 0.9871 | 0.9944 | 0.9907 | 0.9984 |
0.0058 | 0.72 | 3000 | 0.0051 | 0.9928 | 0.9930 | 0.9929 | 0.9988 |
0.0058 | 0.79 | 3300 | 0.0036 | 0.9969 | 0.9920 | 0.9945 | 0.9991 |
0.0048 | 0.86 | 3600 | 0.0047 | 0.9930 | 0.9947 | 0.9938 | 0.9990 |
0.0048 | 0.93 | 3900 | 0.0053 | 0.9863 | 0.9965 | 0.9914 | 0.9985 |
0.0052 | 1.0 | 4200 | 0.0033 | 0.9985 | 0.9909 | 0.9947 | 0.9991 |
0.0029 | 1.07 | 4500 | 0.0039 | 0.9938 | 0.9954 | 0.9946 | 0.9991 |
0.0029 | 1.14 | 4800 | 0.0038 | 0.9981 | 0.9906 | 0.9943 | 0.9991 |
0.0034 | 1.22 | 5100 | 0.0044 | 0.9937 | 0.9934 | 0.9936 | 0.9989 |
0.0034 | 1.29 | 5400 | 0.0040 | 0.9884 | 0.9959 | 0.9921 | 0.9987 |
0.0027 | 1.36 | 5700 | 0.0040 | 0.9975 | 0.9910 | 0.9942 | 0.9990 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1+cu116
- Datasets 2.9.0
- Tokenizers 0.13.2
- Downloads last month
- 3
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.