1.3b-handwritten-v1-after-book
This model is a fine-tuned version of /models/1.3b-dalio-principles-book on the AlekseyKorshuk/dalio-handwritten-io dataset. It achieves the following results on the evaluation set:
- Loss: 2.0566
- Accuracy: 0.0669
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: 3e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- total_train_batch_size: 32
- total_eval_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- num_epochs: 3.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 2.3721 | 0.2 | 1 | 2.2148 | 0.0641 |
| 2.241 | 0.4 | 2 | 2.2148 | 0.0641 |
| 2.469 | 0.6 | 3 | 2.1348 | 0.0653 |
| 2.3735 | 0.8 | 4 | 2.1309 | 0.0648 |
| 2.2755 | 1.0 | 5 | 2.1133 | 0.0652 |
| 2.0428 | 1.2 | 6 | 2.0938 | 0.0659 |
| 1.764 | 1.4 | 7 | 2.0781 | 0.0659 |
| 1.7458 | 1.6 | 8 | 2.0781 | 0.0661 |
| 1.868 | 1.8 | 9 | 2.0820 | 0.0660 |
| 1.9548 | 2.0 | 10 | 2.0703 | 0.0663 |
| 1.6772 | 2.2 | 11 | 2.0605 | 0.0665 |
| 1.3997 | 2.4 | 12 | 2.0566 | 0.0668 |
| 1.3717 | 2.6 | 13 | 2.0547 | 0.0669 |
| 1.5284 | 2.8 | 14 | 2.0547 | 0.0667 |
| 1.2264 | 3.0 | 15 | 2.0566 | 0.0669 |
Framework versions
- Transformers 4.25.0.dev0
- Pytorch 1.12.1+cu113
- Datasets 2.3.2
- Tokenizers 0.12.1