train_run-SmolLM2-135M-Instruct-fadel-full-arabic-diacritization-bs16
This model is a fine-tuned version of HuggingFaceTB/SmolLM2-135M-Instruct on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0463
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: 0.0002
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Use OptimizerNames.PAGED_ADAMW with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: constant
- lr_scheduler_warmup_ratio: 0.03
- num_epochs: 1
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.1347 | 0.0623 | 500 | 0.1323 |
| 0.0996 | 0.1247 | 1000 | 0.1021 |
| 0.0808 | 0.1870 | 1500 | 0.0867 |
| 0.0704 | 0.2493 | 2000 | 0.0773 |
| 0.0661 | 0.3117 | 2500 | 0.0693 |
| 0.065 | 0.3740 | 3000 | 0.0658 |
| 0.0606 | 0.4364 | 3500 | 0.0647 |
| 0.0586 | 0.4987 | 4000 | 0.0589 |
| 0.0564 | 0.5610 | 4500 | 0.0557 |
| 0.0544 | 0.6234 | 5000 | 0.0544 |
| 0.0499 | 0.6857 | 5500 | 0.0526 |
| 0.0497 | 0.7480 | 6000 | 0.0517 |
| 0.0464 | 0.8104 | 6500 | 0.0491 |
| 0.0477 | 0.8727 | 7000 | 0.0487 |
| 0.0423 | 0.9350 | 7500 | 0.0483 |
| 0.0472 | 0.9974 | 8000 | 0.0463 |
Framework versions
- PEFT 0.14.0
- Transformers 4.51.1
- Pytorch 2.5.1+cu124
- Datasets 3.5.0
- Tokenizers 0.21.0