smolm-mlm-bpe-unmask-seed_999
This model is a fine-tuned version of models/smolm-mlm/config.json on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.7174
- Accuracy: 0.4453
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: 0.0005
- train_batch_size: 64
- eval_batch_size: 512
- seed: 999
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 24000
- num_epochs: 10.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 3.561 | 1.0 | 11938 | 3.5141 | 0.3483 |
| 3.3506 | 2.0 | 23876 | 3.3628 | 0.3569 |
| 3.1725 | 3.0 | 35814 | 3.1874 | 0.3836 |
| 3.0509 | 4.0 | 47752 | 3.0693 | 0.3987 |
| 2.9538 | 5.0 | 59690 | 2.9730 | 0.4077 |
| 2.8792 | 6.0 | 71628 | 2.8898 | 0.4234 |
| 2.8026 | 7.0 | 83566 | 2.8437 | 0.4270 |
| 2.7334 | 8.0 | 95504 | 2.7835 | 0.4361 |
| 2.6586 | 9.0 | 107442 | 2.7487 | 0.4434 |
| 2.6413 | 10.0 | 119380 | 2.7041 | 0.4486 |
Framework versions
- Transformers 4.32.1
- Pytorch 1.13.1+cu117
- Datasets 2.12.0
- Tokenizers 0.13.3