swinv2_arocr_tiny_encoder
This model is a fine-tuned version of /lustre07/scratch/gagan30/arocr/models/swinv2_arocr_tiny/config.json on the /lustre07/scratch/gagan30/arocr/Hindawi dataset. It achieves the following results on the evaluation set:
- Loss: 0.0519
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: 1337
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10.0
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.0891 | 1.0 | 8078 | 0.0628 |
| 0.0465 | 2.0 | 16156 | 0.0595 |
| 0.0639 | 3.0 | 24234 | 0.0570 |
| 0.0608 | 4.0 | 32312 | 0.0548 |
| 0.0487 | 5.0 | 40390 | 0.0554 |
| 0.059 | 6.0 | 48468 | 0.0533 |
| 0.0677 | 7.0 | 56546 | 0.0525 |
| 0.0555 | 8.0 | 64624 | 0.0521 |
| 0.0502 | 9.0 | 72702 | 0.0520 |
| 0.0496 | 10.0 | 80780 | 0.0519 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.12.0
- Datasets 2.7.1
- Tokenizers 0.11.6
- Downloads last month
- 8
Inference API (serverless) does not yet support transformers models for this pipeline type.