<!-- 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. -->
layoutlmv3-finetuned-DocLayNet
This model is a fine-tuned version of microsoft/layoutlmv3-base on the doc_lay_net-small dataset. It achieves the following results on the evaluation set:
- Loss: 0.5644
- Precision: 0.6179
- Recall: 0.7238
- F1: 0.6667
- Accuracy: 0.8720
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: 1e-05
- train_batch_size: 2
- eval_batch_size: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- training_steps: 1000
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 1.3383 | 0.58 | 200 | 0.8358 | 0.3007 | 0.4381 | 0.3566 | 0.7724 |
| 0.8308 | 1.16 | 400 | 0.6735 | 0.4634 | 0.5429 | 0.5 | 0.8084 |
| 0.518 | 1.74 | 600 | 0.5706 | 0.5373 | 0.6857 | 0.6025 | 0.8399 |
| 0.3856 | 2.33 | 800 | 0.6303 | 0.6032 | 0.7238 | 0.6580 | 0.8648 |
| 0.2558 | 2.91 | 1000 | 0.5644 | 0.6179 | 0.7238 | 0.6667 | 0.8720 |
Framework versions
- Transformers 4.27.3
- Pytorch 1.13.1+cu116
- Datasets 2.10.1
- Tokenizers 0.13.2
How to Train & Inference:
Check this out this repo: https://github.com/mit1280/Document-AI
