metadata
license: apache-2.0
base_model: riotu-lab/ArabianGPT-03B
tags:
- generated_from_trainer
model-index:
- name: model_outputs
results: []
model_outputs
This model is a fine-tuned version of riotu-lab/ArabianGPT-03B on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 5.9804
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: 0.0003
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- num_epochs: 4
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 6.1892 | 0.26 | 10 | 4.8197 |
| 4.7341 | 0.51 | 20 | 4.1928 |
| 4.1782 | 0.77 | 30 | 4.0742 |
| 4.1255 | 1.03 | 40 | 4.0078 |
| 3.3704 | 1.28 | 50 | 4.1840 |
| 3.4364 | 1.54 | 60 | 4.4455 |
| 3.3383 | 1.79 | 70 | 4.3615 |
| 3.1346 | 2.05 | 80 | 4.3538 |
| 2.451 | 2.31 | 90 | 5.5225 |
| 2.456 | 2.56 | 100 | 5.2477 |
| 2.3692 | 2.82 | 110 | 5.0649 |
| 2.2492 | 3.08 | 120 | 5.3660 |
| 1.8112 | 3.33 | 130 | 5.8835 |
| 1.8044 | 3.59 | 140 | 5.9958 |
| 1.8347 | 3.85 | 150 | 5.9804 |
Framework versions
- Transformers 4.39.3
- Pytorch 2.1.2
- Datasets 2.18.0
- Tokenizers 0.15.2