|
--- |
|
language: |
|
- ar |
|
library_name: peft |
|
tags: |
|
- generated_from_trainer |
|
datasets: |
|
- dalyaff/darebah-2K |
|
base_model: microsoft/phi-2 |
|
model-index: |
|
- name: phi-2 |
|
results: [] |
|
--- |
|
|
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You |
|
should probably proofread and complete it, then remove this comment. --> |
|
|
|
# phi-2 |
|
|
|
This model is a fine-tuned version of [microsoftl](https://huggingface.co/microsoftl) on the dalyaff/darebah-2K dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.5861 |
|
|
|
## 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: 2.5e-05 |
|
- train_batch_size: 2 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 4 |
|
- total_train_batch_size: 8 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 5 |
|
- training_steps: 2500 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 1.2333 | 0.44 | 100 | 1.1475 | |
|
| 1.0155 | 0.89 | 200 | 0.9995 | |
|
| 0.9331 | 1.33 | 300 | 0.9305 | |
|
| 0.8308 | 1.78 | 400 | 0.8755 | |
|
| 0.7955 | 2.22 | 500 | 0.8337 | |
|
| 0.7788 | 2.67 | 600 | 0.7974 | |
|
| 0.7205 | 3.11 | 700 | 0.7685 | |
|
| 0.7588 | 3.56 | 800 | 0.7405 | |
|
| 0.7154 | 4.0 | 900 | 0.7193 | |
|
| 0.6913 | 4.44 | 1000 | 0.7022 | |
|
| 0.6049 | 4.89 | 1100 | 0.6864 | |
|
| 0.6546 | 5.33 | 1200 | 0.6672 | |
|
| 0.6406 | 5.78 | 1300 | 0.6576 | |
|
| 0.6237 | 6.22 | 1400 | 0.6433 | |
|
| 0.6016 | 6.67 | 1500 | 0.6313 | |
|
| 0.5746 | 7.11 | 1600 | 0.6263 | |
|
| 0.5786 | 7.56 | 1700 | 0.6156 | |
|
| 0.592 | 8.0 | 1800 | 0.6087 | |
|
| 0.5908 | 8.44 | 1900 | 0.6030 | |
|
| 0.598 | 8.89 | 2000 | 0.5980 | |
|
| 0.5998 | 9.33 | 2100 | 0.5937 | |
|
| 0.5859 | 9.78 | 2200 | 0.5909 | |
|
| 0.5767 | 10.22 | 2300 | 0.5888 | |
|
| 0.5786 | 10.67 | 2400 | 0.5872 | |
|
| 0.5911 | 11.11 | 2500 | 0.5861 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.7.2.dev0 |
|
- Transformers 4.38.0.dev0 |
|
- Pytorch 2.1.0+cu121 |
|
- Datasets 2.16.1 |
|
- Tokenizers 0.15.1 |
