---
tags:
- generated_from_trainer
model-index:
- name: arzwiki_mlm
  results: []
metrics:
- perplexity
license: mit
datasets:
- SaiedAlshahrani/Egyptian_Arabic_Wikipedia_20230101
language:
- ar
library_name: transformers
pipeline_tag: fill-mask
widget:
- text: الهدف من الحياة هو  <mask>
---

<!-- 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. -->

# arzwiki_mlm (arzRoBERTa)

This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset.
It achieves the following results on the evaluation set:

- Pseudo-Perplexity: 

## 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: 0.0001
- train_batch_size: 256
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.98) and epsilon=1e-06
- lr_scheduler_type: linear
- num_epochs: 5


### Training results

| Epoch | Step  | Training Loss | 
|:-----:|:-----:|:-------------:|
| 1     | 2500  | 2.038300      | 
| 2     | 5000  | 0.878800      | 
| 3     | 7500  | 0.682800      | 
| 4     | 10000 | 0.613100      |
| 5     | 12500 | 0.574500      | 

| Train Runtime  | Train Samples Per Second | Train Steps Per Second | Total Flos                | Train Loss |  Epoch   |
|:--------------:|:------------------------:|:----------------------:|:-------------------------:|:----------:|:--------:|
| 14677.117400   | 248.119000               | 0.970000               | 120746231839334400.000000 | 0.908513   | 5.000000 |



### Framework versions

- Datasets 2.9.0
- Tokenizers 0.12.1
- Transformers 4.24.0
- Pytorch 1.12.1+cu116