table-transformer-finetuned
This model is a fine-tuned version of microsoft/table-structure-recognition-v1.1-all on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.2890
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-06
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 8
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 20
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 9.2523 | 1.0 | 14 | 9.3207 |
| 7.9949 | 2.0 | 28 | 9.0533 |
| 8.2292 | 3.0 | 42 | 8.6547 |
| 7.1495 | 4.0 | 56 | 8.0844 |
| 6.9055 | 5.0 | 70 | 7.3339 |
| 7.0771 | 6.0 | 84 | 7.2344 |
| 5.6554 | 7.0 | 98 | 6.6026 |
| 5.5633 | 8.0 | 112 | 6.3610 |
| 4.9506 | 9.0 | 126 | 5.9567 |
| 4.0856 | 10.0 | 140 | 5.9191 |
| 4.4286 | 11.0 | 154 | 5.6374 |
| 2.9043 | 12.0 | 168 | 4.9525 |
| 3.4755 | 13.0 | 182 | 4.6846 |
| 3.4152 | 14.0 | 196 | 4.3171 |
| 2.8456 | 15.0 | 210 | 3.5661 |
| 2.4149 | 16.0 | 224 | 3.6150 |
| 1.9328 | 17.0 | 238 | 3.2264 |
| 1.8503 | 18.0 | 252 | 2.8540 |
| 1.596 | 18.5981 | 260 | 2.2890 |
Framework versions
- Transformers 4.48.3
- Pytorch 2.5.1+cu124
- Datasets 3.3.2
- Tokenizers 0.21.0
- Downloads last month
- 88
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support